Junk chess

i.e. the kind of chess that makes you form bad habits. What form can it take, and how can you reform good habits afterwards?

Time controls shorter than rapid are some of the obvious answers. Rapid time controls might also be included, but as a rapid player myself, I can't tell if that's a part of my problem or not.

Also, how bad do your openings need to be in order for you to conclude that you might be playing junk chess based on them alone? I fear that it might be my problem. I've been shying away from "mainstream" openings for a long time now (for fears of getting out-theorised), and I've been almost exclusively playing such openings as 1. Nc3 (into Van Geet, Jobava London, or Veresov), 1. d3 (into White Lion, Old Indian Attack, or KIA), the 3... Qd8 Scandinavian, the Philidor Lion, and various transpositions after 1... d6.

I've invested much time and effort into mastering these through practical games, but even though said openings aren't by any means refuted and I'm not a very high-level player (rapid rating around 2000 on Lichess), I already feel like it might not be good for me to play those. I don't know what to do. Am I playing junk chess or not?

The reason why classical openings are played is because they make sense. My suggestion would be to do a quick-starter courses on chessable for some solid opening choices - there is no cost involved (I think you can do certain amount of quick-starters without any cost). As white Jobava and London are quite a reasonable choice, as black Caro-Kann against e4 and perhaps classical Slav for d4 or simply QGD would be fine. The point is that even if you don't know theory 20 moves in, because these openings make sense they will not leave you in an overwhelmingly worse position. If you like playing fianchetto as black, go for Pirc and KID though they are a bit tougher to master. Since you play Jobava as white you are probably aware of the simple plans against KID so that might help you in defending against it.

Junk chess is relying on traps and tricks which often works well in blitz but not as well in longer time controls. However, if you lay reasonable openings you can use blitz as a way of flushing out some middle-game plans and learning some patters. The most important thing you need to learn is to calculate - you can't just trust your instincts in longer time controls and faster time controls force you to do the exact opposite. So, my advice to you is to spend a bit of time on opening theory of some classic openings, you can use blitz as a way of practicing the openings as you are learning them but don't expect anything more out of blitz. It doesn't translate beyond that to slower time controls.

The best chess players in the world (Carlsen, Nakamura, others) all regularly state that getting your opponent out of book is one of the most important things you can do in modern chess... the openings you've mentioned are not shallow trappy openings, they're complex and rare, which is really what you want these days... I don't think your concerns are very justified. 
The critical question that'd need to be asked here is whether you know the opening well but you're being out moved ordered consistently by your opponents, or whether you just don't know the branches of the opening / aren't playing it correctly / whether your problem has nothing to do with the opening. And based on your elo and the openings you've listed I know the answer to that is the latter.
At least half the benefit of getting your opponent out of book is getting them into your book where you can punish their mistakes... but if you don't really know the book well, including how to punish the different common mistakes your opponent can make, you're not capitalizing on the opportunities of this approach. Instead you're just intuitively playing a position that's relatively inferior to mainline openings... and if at your elo your opponents don't know the theory of the mainline openings anyway... you're just left at a relative disadvantage by playing the obscure openings, there's no real benefit of them. 
If you're playing these openings for traps, rather than for positional advantages, again I think you're not playing it right. 
If you're not able to play positional chess well enough to capitalize on these advantages... practice positional chess. 
Usually I recommend for people at lower elo not to fixate on a specific opening... you ultimately need to know alot of openings because they can transpose into one another. And learning many different openings is a good way to improve your positional / tactical chess. For example, if you struggle with positional chess you should try playing the queens gambit declined, since that opening is all about the classical chess principles and positional play. You could consider playing this for a while until you reach a higher elo / develop positional skill and where getting your opponent out of their book will have more benefit and you can capitalize on it. 
But the point is it's not the junk opening or something junk about the approach to chess, it's you playing like junk (no offense). 
For openings like the Pirc / Modern / Veresov you won't run into problems with people even knowing the main lines up until like 2000 rapid (chess.com elo, not lichess). And even then they still won't know the openings very deeply. Meanwhile there are always ways to equalize in those lines (like in almost all lines), and plenty of novelties all throughout them. There are grandmasters that play all the openings you've mentioned, including some that play them as their main repertoire (Jabova London for example).
Though I'm not as big a fan of the Van Geet, considering you have other better options... it's an epic opening, don't get me wrong, but the opponent can transpose that into a sicilian and play it pretty intuitively and be fine, also e5 players can transpose it into a vienna which they will be familiar with probably... I'd rather play the Trompowsky with the jobava or veresov, at least at your level... thsi will not putso much of a burden on you to learn theory. 
And the Pirc / Modern specifically is difficult for beginners to play... I wouldn't really recommend that. 
Jabova / Veresov are good ones though. I'd also recommend the Trompowsky to compliment these.
But anyway - mess around with all the major systems for a while - kings indian setup, slav setup, QID/nimzo, queens gambit (should probably be your first), ruy lopez, italian, various sicilians, french defense, caro-kann ... just learn the beginning moves and play the positions. 
For example there are many transpositions from the chigorin into the french defense / pirc, and transpositions from the Trompowsky into the queens gambit.

Keep yourself informed about the latest developments in your chosen openings. Even unconventional openings can benefit from a solid understanding of the underlying principles.

Thank you for the kind and helpful words, everybody! There was a lot written in just a few comments during my recent absence, so I'mma boil my answer to all of them down to a few concise bullet points:

- I don't play any of my openings for traps, but it is a combination of the other mentioned factors: I don't know all the branches (because a number of them gets played rarely and/or badly), I get "outmoveordered" (although I practice my move orders a lot with an analysis board), and I also lack the understanding of the opening sometimes (because I don't play mainstream openings, so my learning resources are scarce);

- a notable drawback to my opening repertoire is that it's built around themes like "always control at least e4 as White/e5 as Black, and then try to also control e5 as White/e4 as Black if you can", which makes me try my hardest to avoid uncomfortable positional features that prevent me from doing that (i. e. when Black plays Bf5) at all costs... but that often gives me a somewhat unnecessarily passive position as a result, and if I were to switch to 1. e4 in the above example, I'd have to contend with the Scandinavian, which doesn't just let me hunker down with d3 and Nd2 next;

- I'm afraid that my calculation (as well as tactical vision) is a real issue, as I've gone too much into "when to calculate" but not "how to calculate" and I find visualisation to be unnecessarily mentally taxing. Back when I was more serious about chess, I calculated and visualised so much that my brain started using too much memory that I needed for other things in my life, so I'm stuck in this limbo between being completely uncaring and going all out, both of which are bad for me if I wanna derive a moderate hobby-level pleasure from chess.

All in all, it seems to me that it's not my openings that are junk, but my mental habits. I'll see what I can do about it.

To really design a good repertoire you have to know what your goals are in chess.
What elo do you want to achieve? 
If you're content in staying below like 2000 fide, and if you don't want to put alot of time in... you should play lines that are much more forcing. i.e. against 1. d4 you could play the albin counter gambit, and not worry about memorizing all that 1. d4 theory. 
If you take your chess more seriously... that won't work. 
So really that's the first question you must answer - what are your goals. 
It's good to design your repertoire to use certain common structures... but I don't think it's worth really handicapping your repertoire to achieve that. It's inevitable that there will be some variance in the structures, and there are other priorities to consider as well. 
Just to illustrate, these are my priorities in a repertoire... this is based on the understanding I'd be content hovering somewhere in the low 2000s, maybe 2300 would be the max I'd aspire to reach... because I have other priorities in life- 
- avoid the really popular mainline systems / get opponent out of book (you've done that well). popularity can be told via lichess analysis by looking at the percentage of games played on lichess
- limit the amount of branches in your opening - (you aren't doing that well. for example, I don't play the nimzo indian or the QID because very early in the opening there are tons of ways white can respond and it explodes in the number of branches. same with the ruy lopez. This sort of opening would be more appropriate if I were intent on becoming a really high level player. The pirc can be quite complex as well, btw). Limiting branches is especially important for casual players. 
- another option for minimizing theory is to play a system opening... i.e. something like the torre attack or the nimzo larsen attack or the kings indian attack or the english fiancetto setup ... this isn't really my style but it works for some people. i.e. you can mostly play the position intuitively and not really have to know exact move orders. But you should know the ideas and different plans that are possible. 
- play openings that work together where possible - ideally they can transpose into one another, but if not go for common structures (you're doing this well but you're taking it too far). But some openings rarely transpose, and lead to obscure positions, they just have other things going for them... like gambits usually don't transpose but because they're relatively forcing or deadly... they're worth playing. 
- should match the type of player you are (i.e. some players are more tactical, for them I'd recommend e4... others are more positional and should probably play d4 or some other opening...
- should be forgiving of mistakes for your side but more sharp for the opponent / punishing of their mistakes. Also the continuations of different offbeat moves should seem natural, and the ideas in the position should be understandable. This is even more important at lower elo... the Pirc is very sharp, if black makes one wrong move he gets punished. It throws people out of book, but you have to know the book yourself. But that's not the kind of player you are, you don't want to deeply study and know the book, including what to do against mistakes of your opponent. 
- openings should have a good winrate at your skill level... I don't think you're doing this well, i.e. you're playing the pirc but you're below master... for this you should use lichess analysis to see the winrates at whatever your target elo level is. For example, of all blacks responses to the queens gambit the albin counter gambit has the best winrate on lichess... for players rated below 2200. 
last thing you should do is create some ... either lichess studies, or even better get chasebase and create some annotated games detailing your repertoire. That allows you to quiz yourself on the repertoire by reviewing the study or annotated game. You can also save this... put it on a thumbdrive if it's a chessbase game, otherwise on lichess it's saved to your account... this is a one-time investment in effort, but after that you can easily remember your repertoire without having to constantly study and review it to remember. So lets say you take a 2 year haitus from chess... you come back, review your studies, and vuala within just a few days you're back up and running. 
- I also think you should feel inspired by the opening... 
So think carefully about all this, and once you make your decisions and feel good about it go ahead and develop your repertoire based on those values, and after that never change it again.
- another thing you should do is download leela chess engine and get it working within some chess interface. I have chessbase, but there are free programs as well... but for learning uncommon openings you can go a long way with leela + youtube videos where you can find them.

I've formatted your points so I can address them in turn, with the notable omission of the ones that use the Pirc Defence as an argument... like, where in this thread have I ever mentioned anything about playing the Pirc? 😆

Now, about your points:

1) I agree about that, and Lichess stats are really nifty, in no small part due to the fact that it can show you the most common opening mistakes, although I rather wouldn't rely on my opponents playing suboptimal moves;

2) Are you sure that I'm not limiting my branches well? Limiting branches is why I don't play 1. e4! 😆 All the other first moves I mentioned for White lead to a somewhat lesser variety of opening positions, in my humble estimate. (I was toying with the idea of playing 1. e4 d5 2. d3 anyway and allowing the queenless middlegame after 2... dxe4 3. dxe4 Qxd1+ 4. Kxd1, as it's close to the Exchange Lion positions in reverse... but it's the alternatives with the passive N(b)d2 that repel me at the moment) I also don't know if you consider the Black Lion branches to be too numerous, but there shouldn't be more than 5 main ones in any case, and in virtually all of them, Black adopts the Old Indian formation which can sometimes become a KID or a Czech Benoni, so those ideas can be carried over, to an extent. Seems pretty limited to me, if I eliminate the Valencian Scandinavian and some Dutch transpos which I only play for variety.

4) Skipping over point no. 3 for the moment, aside from the example I've given in my answer to your point no. 2 + playing positions with colours reversed, there are other openings I play only against certain lines, and they all synergise well, for example: after 1. d4 d6 2. Bf4, I pivot back to 2... g6 and I prepare to play ...e5 in a couple of moves, which is very effective against the London setup; also, after 1. e4 d6 2. Nc3, with the intention of 3. f4, I can pivot to a Modern or even a Hippo again, and the Gurgenidze System can also help me against 2. f4. That should be a good example of openings working together.

3) Back to this, I do employ system openings as much as I can, but as I've said before, 1... d5 + 2... Bf5 is annoying and difficult to stop in most non-e4 openings, barring some gambits that might not be a good long-term investment, so my best chance might be just playing 1. d4, 2. e3, and trying to exchange the Bishops immediately with 3. Bd3, then optimally achieving a Colle-Zukertort System or even a Stonewall Attack formation. My impressions flip back and forth, though. I also like the Vienna-esque fianchetto system with e4, Nc3, g3 and Bg2, but again - it's an e4-system, and it cannot be played against everything.

5) As someone constantly torn between a deep desire to play like Tal and an equally deep desire to play like Petrosian (with Alekhine being another favourite who regrettably didn't play most of the openings I do, or at least not in a modern enough way), I have to conclude that I'm not a strong enough player to have a style. If sucking at tactics (especially defensive ones) and endgames makes me a "positional attacker" or whatever, then so be it. 😆 I will say this: I think I do prefer closed games to open games, and I like having long-term advantages and harmonious pawn structures, even though I'll angle towards a direct attack whenever possible. The closest I might go towards chaotic positions are the Open Sicilians with all their glorious thematic sacrifices, but hey - I'd need to get to those positions first!

6) Any particular reason you find Lc0 preferable to the built-in Stockfish engines that Lichess uses?

The Black Lion can be reached via the Philidor or the Pirc defense.... people who use it as a main feature of their repertoire typically reach it via the Pirc via 1. d4 d6 2. e4 Nf6 
In any case, the discussion points regarding the Pirc apply to the Black Lion variation regardless of how you're reaching it.
Were you not able to connect the dots there, or are you just getting an attitude because you don't like the suggestion...? 
Be thankful a better player than you has taken the time to type out a detailed response giving you advice, I don't have to do this. 
Again, in hyper modern openings... and the Pirc, or the Black Lion, or whatever you want to call it, is a hypermodern opening... you cede the center. Therefor you don't apply pressure to your opponent, your opponent is free to play any move they like. As a result there are many, many branches in those openings based on every potential move your opponent can make. What you appear to be doing is "limiting the branches" by simply not studying your opponents "mistakes"... That's not a feature of the opening it's a feature of your approach to chess...
There are also different degrees of mistakes, some that can be punished immediately and forcefully, some positional and subtle... in hyper modern openings the opponents mistakes typically are positional and subtle. 
You claim you don't want to rely on your opponent making mistakes... but that's one thing you can always rely on in chess, and doubly so in a hyper modern opening that isn't common... If you don't like that... I'd recommend you play a system opening. A hyper modern defense is all about punishing the opponents mistakes, their mistake is your opportunity to break open the center, and if you miss that opportunity you will probably lose.
And it's usually a very small window of opportunity you have. These defenses are super-sharp, and usually way more punishing of blacks mistakes than of whites... which can be a mistake in responding to whites mistakes. If you don't want to account for that... and it takes alot of time and study of theory to do that ... I suggest playing something else.
The very idea of playing a sub-optimal opening is to induce mistakes by getting your opponent out of book. To assume your opponent won't make mistakes, and to be unconcerned with punishing them... is to assume that you'll end up in a relatively worse position. 
But you don't have to take my advice, you are free to continue doing exactly what you're doing. Of course this isn't just my advice, it's a common thought on hypermodern defenses that's been expressed by many great players.

Openings that limit the number of possible branches are those that apply pressure in a way that restricts the opponents options, and where mistakes have a direct refutation... 1. e4 and 1. d4 control the center from move 1, and in doing so they limit the opponents possible replies. 1. d4 probably reduces the potential valid replies the most, in my experience... 1. Nc3? Not so much - your opponent can play pretty much all the 1. e4 defenses... the caro-kann, the sicilian, the french, the vienna game, the pirc. If anything you limit the range of your potential responses to these defenses... i.e. you have to play the classical karo-kann, the paulsen french, the vienna game, the closed sicilian, etc.. In addition your opponent can play 1... d5 and then you're playing all the chigorin sidelines. Or a line in the scandinavian, which again you could play via 1. e4 if you wanted. So... how have you limited the opponents options by avoiding 1. e4 here, exactly...? You've actually just limited your options. These lines can all be reliably reached from 1. e4, they're 2nd move choices from white.
Granted you can play a sideline and avoid transposing, but you're not really reducing the amount of complexity and potential moves, just playing weird moves to avoid common, established theory. But to play well you'll still need to know the line...
1. d4 > 2. Nc3, in contrast, prevents the sicilian and the vienna game. Your opponent can play an old benoni, but that position honestly is not very good, not many people play it and you don't need to know very much to respond, just play d5 and play Bf4, rotate your kings knight to c4 ... the sicilian is huge and alot harder to navigate. And the englund gambit? It will take you about 15 minutes to learn to refute it - nothing like say... the scotch or the vienna gambit or the 4 knights spanish, which there are whole books about.
And what do you play against 1... d5? Let's say you're playing the veresov or the jabova. Would I rather my opponent get to play their favorite e4 line of their choosing, or settle for the old benoni / face my jabova or veresov that they've never seen? Your opening actually seems to give them an opportunity to escape the main weapon you have prepared. 
Again, if you're hung up on 1... d5 2... Bf5 - a sideline - in the chigorin this is played 6% of the time, in queens gambit it's played 1% of the time - your problem could be greatly reduced by just playing a regular move order. Or develop a specific line against it - this isn't a critical line, it's a sideline, it's not a real problem... there's just a range of solutions that you have to weigh the pros and cons of then choose from.
So for this reason... I'd say 1. Nc3 is fun, full of great lines, it's a cool way to play / confuse lower elo players, but ultimately it's relying on your opponent making a transpositional mistake while giving a better player an opportunity to escape your main weapon and either play a line they're more comfortable with, or making you play a sideline...
If you love the lines, and just really want to play them all, it could still make sense for you - I think the lines are great - but if your goal is to spend not alot of time on chess.... or if your goal is to reach very high elo... it's not a move that helps you with that.

Lc0 sees moves stockfish and other engines just do not see... for example, the greco variation of the QGA was written off as garbage until Lc0 came out and found a continuation to a move all the engines had overlooked, now the greco variation is among the most popular responses to 3. e4 in the QGA. 
If you're exploring unusual defenses... you don't want to miss critical moves. Or critical refutations.
I can tell you I've explored some rare moves and leela finds moves that have never been played on lichess, my suspicion is the other engines just did not find the move. 
Also, you'd rather the suggested moves be playable from a human standpoint, and the way leela works using a neural network... more closely emulates human thought processes, including intuition, than a traditional engine based on a more traditional algorithm.

1. Wait, are we talking about the same Black Lion?

I get that the Lion is only reliably arrived upon via a move order which is used to reach the Pirc Defence without transposing from the Modern, but that does not automatically make every opening reached via that move order a hypermodern one. Since ...e5 can practically be played on either the 3rd or the 4th move in all of the variations I've presented, sans the 2. Nf3 + 3. e5 one, it's resoundingly all still a part of the Philidor Defence, which is very much a classical opening because the early ...e5 stakes a claim in the centre. The Lion structure is more similar to the Spanish and Italian ones than to the Pirc one, because in the Pirc, the combination of ...Nf6 and ...g6 is mandatory but ...e5 isn't, just like how ...g6 is not mandatory in the Philidor but ...e5 very much is.

By definition, hypermodern openings strive to control the centre from afar (by using pieces instead of pawns), allowing the opponent to establish more pawns in the wider centre (if he so desires) in order to attack them afterwards, right? Well, all the Philidor variations, except the Lion with 3... Nbd7 4. f4 (which Black can circumvent a move earlier while keeping it a Lion), forcibly keep White's pawn presence in the centre at bay, at least until the middlegame.

All things considered, I see no need for anyone to get all worked up and go all "respect your superiors, whippersnapper" on me over something that's fundamentally a semantic misunderstanding, at least as far as I can tell. There was nothing inherently dismissive about my remark, but if I'm wrong about the aforementioned structural similarities, please let me in on your perspective on the matter.

2. The possible dispute over the definition of hypermodern notwithstanding, I get what you mean by branches now - opportunities for the opponent to deviate from one exact plan/formation on his behalf and still get a relatively good position; well, I suppose that's the price of playing openings which mostly stick to a single structure on the other side. The openings I strive to use don't prevent my opponents from playing the way they want, but they do prevent them from preventing me from playing the way I want.

When I pick any opening, I wanna feel comfortable playing all variations that could arise from it, but if there's even one variation that my opponents could figure out annoys me (and/or otherwise diminishes my winning chances by a sizeable enough margin) and start playing it against me almost exclusively, then I switch to anotger opening if I can help it; and if I cannot, then I work extra hard to combat those positions which challenge me the most. I like all the vast transpo tricks in the Van Geet/Dunst, but if I can't regularly find my way through either the main line or the Jobava London/Veresov Attack, then 1... d5 is enough of a problem for me to have to stop... which I did. Dunno if that's the right mentality to have, but it is how I've been playing so far.

3. Actually, this point really made me think! I should really go back to analysing openings without relying on the engine evaluations so heavily, it feels like it's been real toxic towards my enthusiasm lately.

4. Very true, even if the Lion isn't technically hypermodern, as I do find that this sentiment applies to it as well. It's more on the "solid but passive" side, but still.

I think this also covers your first follow-up comment, and as for the Lc0 explanation - noted, I'll keep it in mind!

The mainline philidor does not play Nf6 on move 3, it plays dxd4 immediately, you'd then followup with Nc6 and you've both resolved some of the central tension and developed a piece to the center, at that point you'd be in a more classical position. Philidor had his own philosophy of chess... he didn't spur a hypermodern revolution but his ideas were certainly a divergence away from classical chess principles, to call the mainline philidor 'resoundingly classical' is nonsense. Anyway, you're not playing a position like that, your pieces are developing to the 7th rank and you're preparing c6.
Firstly, let's not skip over and ignore the Pirc, the first 3 moves matter - if you're entering the black lion via the pirc - 1. d4 d6 2. e4 Nf6 3. Nc3 e5
If the opponent plays any other moves you'll remain without a center pawn for a while, i.e. 2. c4 and you're in a kings indian, 2. Nf3 or 2. Bf4 looks like you also can't push your pawn quickly. So you're playing hypermodern defenses here. Even if you did make a convincing case that the black lion isn't hypermodern, it would be based on a very superficial concept of the opening, i.e. the number move on which I play e5 - meanwhile all the essential characteristics of a hypermodern opening are present - your pieces are not free to develop, your bishop / queen / knight are all stuck behind your pieces all backed up on the 7th rank and the central tension isn't resolved... white is free to push d5 if he pleases, and you'll have to undermine that. for example, in the mainline pirc you could play a delayed e5 and still transpose to the same position... you'll be timing some pawn push to resolve the tension. OR you'll be counting on your opponent to resolve it via dxe5 dxe5 Qxd8, but that's whites decision to do... it's a semantic debate that someone else can squabble over, we want to think about how to minimize branching, so for our purposes what really matters is you're not able to restrict your opponents options, because your pieces are developing to the 7th rank and you have only limited central control. 
For example, don't focus on the main line, lets just take a look at a sideline white can play-
1. d4 d6 2. e4 Nf6 3. Nc3 e5 4. d5
Blacks most common response here is 4... Be7. Now go on lichess and take a look at the number of 5th moves in this position for white. 
There are 7 different moves that are played over 7% of the time, and 2 more that are played... relatively frequently, 4% of the time. 
So you see, the opponent is free to make pretty much any move because you just played Be7, and that doesn't apply force to restrict the opponents movement. That is my point, you say you don't want to put the time and effort into learning theory, but when you play openings where you do not exert pressure... you end up needing to learn alot more theory. 
We're just looking at 5th moves here, but at 1800 level or so, when you develop a serious repertoire... you probably ending up knowing the opening 8-12 moves deep or so. 
In contrast, there are openings such as the albin which are much more forcing, where there are 1-2 valid moves and anything else is easily punished... do you see the difference we're talking about? What kind of opening do you think a player who wants to put minimal effort into their chess would be successful in? 
All that said... looking at the lichess stats the black lion does have a pretty amazing winrate.... because alot of white players just immediately dxe5 and queen trade, and White players obviously just have no clue how to respond to it... 
Based on those stats I'd say the line can probably work anyway, despite the branching issue... though I still think this is an opening that requires effort to be good at. But the deeper problem remains, because you still have to reach it, and you have to play all these other hypermodern openings like the KID, which is extremely theoretical and complex. 
You don't have to take my advice, but my advice to someone not looking to spend alot of time on chess would be to play something very forcing or dynamic like the albin, the blackmar diemer gambit, etc.. But I do think the black lion looks very cool... and if you actually are going to spend alot of time working on the opening then great, more power to you.

1) Okay, fair enough, I definitely get your point now. Never thought of Philidor's ideas as even proto-hypermodern, if only because he insisted on pawn play so much. Also, it's weird to me that the Exchange Variation (of the Philidor, not of the Lion) is technically still considered the main line, but that's probably besides the point as well.

2) Aside from the fact that I'd rather go 4... g6 or even 4... c5 in that line, the more we talk, the more I get the impression that we are, or should be, discussing the issue of understanding vs. memorisation.

How should I go about sharing my thoughts on the matter? The way I see it, how much theory I need to learn depends on how critical and/or (subjectively?) positionally unintuitive the lines are. For example, one of the reasons why I play 5... h6 instead of 5... Be7 in the Lion's Claw (the main line I provided a couple of comments ago) is that after 6. dxe5 dxe5 7. Bxf7+ Kxf7 8. Nxe5+ Kg8 9. Ng6 Rh7 in the former, the moves I need to play are much clearer and easier for me to understand than after 6. Ng5 O-O 7. Bxf7+ Rxf7 8. Ne6 Qe8 9. Nxc7 Qd8 10. Nxa8 and either 10... b5 or 10... b6 in the latter. Both of these Lion's Roar variations are no-brainers up to these early points, and both leave Black standing better afterwards, but it's much more simple and exact for me to exploit and convert Black's advantages in the first one.

It is true, admittedly, that in the Lion, especially in the main lines with ...g5 (without Black castling kingside), Black needs to remain vigilant about the possible sacrifices (Bxf7+, Nxe5, Bxg5 etc.) before he move-orders his way "out of the woods", and that's what makes it especially, even uniquely challenging; however, for the sake of argument, let's consider another opening: the Czech Benoni. (I really do mention it at least every other comment, don't I? 😆)

White's got a permanent space advantage, and there are many possible setups and move orders available to him; however, is that really automatically much more theory to learn for Black? Since the centre is closed, it's all about manoeuvring, and practically nothing is forced. The name of the game is understanding piece placements, trades, and pawn breaks. A space advantage is the easiest to have but the hardest to convert, so at all but the highest levels, it's much more likely for Black to make tactical blunders in all other positions than to get squeezed to death in this one.

This might be down to whether you consider a move to be theoretical if you can always find it and reliably play it based on your understanding. We learn theory so that we can be prepared for what we cannot figure out on our own within a reasonable amount of time, no?

3) Tying in with the previous point, the irony is that White players should at the very least be able to play natural, normal-looking, developing moves, given that Black's position is already slightly passive... and yet, they don't. It's not the simplest discussion one could imagine, as one obviously has to take into account such factors as average White ratings, the psychology of getting surprised by something one just simply hasn't faced before (even if it merely *does not look dangerous*, regardless of whether it is or isn't in actuality). That, in turn, begs the question: why should I be concerned with my opponents having a multitude of viable options, if that trips up their decision-making more than mine? The answer, of course, is that, once they get past their unfamiliarities at a certain level, they get to play natural moves, and I have to work harder to punish them for those... bit I think that's one of the psychological advantages of playing the Black Lion: if you (as White) play those "natural" moves on autopilot, without putting in any sort of a deliberate effort to understand the subtle differences between the various options at your disposal, then you yourself are at a high risk of being subjected to an autopilot attack as well.

4) Just for the sake of example, since you mentioned the Blackmar-Diemer Gambit: even though it might be useful for "solving" both my Bf5 problem and my Scandinavian problem, not only do I need to learn at least a Caro-Kann line and a French line in case Black decides to decline the gambit, but also, in at least a couple of lines which are advocated for Black if he accepts the gambit, White needs to work especially hard in order to compensate for the lost pawn. When everything is very tactical, everything is very exact, but in this case, it's not completely exact for both sides like it is in some pure minefield openings such as the Traxler Counterattack. Black maintains his easy, intuitive solidity, and it's up to White to prove he can instill chaos in the ensuing positions and outcalculate Black.

Maybe this is the "positional attacker" speaking from within me, but I've always found playing such openings, including the Smith-Morra Gambit, to be a chore. I can handle forcing queenless middlegames in order to reach equality, but I can't handle forcing my way through long tactical sequences as effortlessly.

A few points regarding learning the opening: 
- If most of the moves in the opening were obvious you'd be much higher rated, but they aren't. chess is a very tricky game, the best moves are usually not obvious. Especially when you're playing a subtle positional opening. Even if the 1st move in a continuation is obvious... again a serious repertoire should be probably 8-12 moves deep, there will inevitably be details deeper down the line that aren't as obvious.
- although you may not remember all the moves of your repertoire, through the process of studying it you still learn common patterns and themes, and these ideas will be more readily available in game if you've familiarized with them before, compared with having never done so.
- I'm not suggesting every move deserves the same amount of attention, some moves are just not good, but in your opening there are alot of moves the opponent can make that aren't immediately and easily punished. I suppose you can assume your opponent will always be just a little more cluess than you... but generally hypermodern(ish) openings are more punishing for blacks mistakes than whites, so if you're both clueless... who loses? 
- at some point you'll reach a level where you're playing good players, they're playing basically valid moves and you'll need every advantage you can get. If it's a common move and not glaringly bad knowing the continuation will simply mean you win more games, that is the bottom line. 
- the problems we're discussing apply not only to the black lion but to the KID and other sidelines you're playing. let's talk about the KID... it is one of the most sharp and theoretical openings. And unlike in the black lion, in the KID you really don't have the benefit of just assuming your opponent won't know the continuations. And black doesn't have the same winrates in the KID that you see in the black lion. 
- most black lion players probably study the opening in depth, I would imagine. Infact... now that I'm looking at the stats, the winrate for black is much higher at higher elo than lower elo... that's probably because black knows the lines better at higher elo, and has just thrown the opponent off.
At higher elo (like 2000-2200ish) knowing the opening at least 8-12 moves deep, including the sidelines, is pretty normal. (I think GMs are usually 12-15 moves deep, and deeper in certain theoretical lines)
Of course alot of how seriously you take this depends on your priorities as a chess player, and that's again up to you to decide. 
*** case in point *** 
Chess: austinwheeler03 vs Shaikidow - 30154766161 - Chess.com

1. You're saying that about my rapid rating of around 2000 on Lichess, right? Honestly, even though I focus on my openings a lot, I wouldn't say that they're the main reason I lose games. I've had many great middlegame positions which I then proceeded to lose due to tactical blunders, especially in time trouble... and that's to say nothing of my troubling lack of endgame technique.

2. Case in point: in this game, I thought I had ...Ng4+ after ...Bxf2+, but by the time I realised White could play Qxg4, it was already too late, so I decided to play Nxe4+ instead, in order to get at least some positional perk afterwards. Not that it really helped me any. I did play that game a bit over two years ago, though, so for my more recent efforts (as well as some older ones), you can find my games on Lichess under this same username. I really don't play seriously on this website.

P. S. Sometimes I eschew obviously good moves due to trying way too hard to avoid annoying moves by my opponent. The ...Bf5 situation is really the cream of the crop in that respect, I've had great troubles with it over a large period of time. If I need to chase it, how? If I need to trade it, when? If I need to accept that it will occupy that diagonal forever, should I just give up and find a way for it never to arrive there in the first place? 😆

Update: I played a couple o' games on Lichess a few minutes ago, and based on how I flubbed some good positions, I can only conclude the following: I AM SLOW. 😅 My mind is just torpid. My sleep schedule hasn't been the best lately, but I don't think that fixing it would help me as much as I'd like it to (even though I should totally do it for myself, regardless of chess). It's not that I don't have a thorough thinking process, it's that I never gave myself a proper chance to optimise it.

I guess I'll have to give classical chess a go if I'm to improve, after all? Or are there other options that could help me improve as well? I've been doing checkmating tactics on Lichess, with the intention of switching to other motifs after I've finished practicing the mates in 4, but it's not like my effort in that department has paid off yet (provided I practice enough, however much that is, which I would also guess I don't 😅).