Setup Openings vs Theory? Which is better for Beginner players?

Many years of playing, losing, and learning, I suppose.

Noam-Bot plays slow and solid, for sure. But he does make tiny mistakes, here and there. They're just less obvious mistakes than the lower-rated bots.

Well, it might then just be semantics ... A player makes a mistake in the opening, then s/he triesto find a better move, then s/he looks up what is usually played in that scenario, but often the move that is recomended by theory is not a checkmate or decisive material gain (what a beginner would understand) ... rather, it might be a move that leads to one type of relatively even game compared to another, or maybe one side has close to a pawn advantage, but neither player at these lower levels can understand how to take advantage of this imbalanced assessment.   So in reality, all a player is learning in the opening is: 1) Don't make this move that leads to a tactic against you (something that is good to learn at this level); and 2) here is what you should play, but you won't understand until you are rated >2000 (which means that the newly learned "good" move is simply memorization by the player).   

I think the OP asks a good question, but a lot of folks want to caricature it into: What should I do with my openings?  A) Have a repertoire that is intensively theoretical and memorize 15 books cover-to-cover --or-- B) Have a repertoire that is simpler and only have to memorize 5 books cover-to-cover.   Of course, there are a lot of beginners that probably do think that way whether they admit it or not.  But if someone wants to know what might be good to play through move 5 ... something that is independent of the opponents' moves (easier, less time needed to devote to it, more time for tactics, endgames, and game score analysis --or-- something that is strategically and tactically rich, but will take much longer to grasp, but better over a 7-10 year haul, then I think that is a reasonable question for someone to ask if they want to "learn correctly."

I feel like I can really agree with your perspective on this. I feel like perhaps I may have mistaken theory for having a large repertoire of uncommon openings. I like learning about gambits and refutations, and also knowing many different lines for various openings such as the queens gambit declined, accepted, The Sicilian Najdorf vs The Dragon Variation, and stuff of the sort. I like to mix things up as long as I'm comfortable getting myself to an even middlegame, or if my opponent makes a mistake I prefer knowing the best move to counter it because of my light studying using the engine/opening database. 

I'm just going to ignore all of the "stick to opening principles" posts here for now as that's not what you asked

I truly believe there's this misconception about 'system' openings at the Beginner level. They always try to label the King's Indian or the London or Colle etc to try to avoid learning theory. The King's Indian is one of the most theory intensive openings in chess, if not the most (Although, by "King's Indian", I reckon you're probably just talking about a Kingside Fianchetto and trying to turtle up). The London system, while certainly not as strict as the King's Indian, still has specific move orders you need to adhere to avoid being blasted off the board with moves like qb6. And imo you can't really play the London system against anything, if you try to play the London vs say the Benoni, or the Dutch, or nf6 + c5, you're just going to end up in a difficult position to play. 

It doesn't hurt to learn opening theory as a beginner, although it's important that you're learning the ideas of the openings, and you're not just memorizing moves. I started learning theory up to move 10 when I was like 900 and never used the deep lines I studied until I was probably 1300-1400ish, but I did learn what types of moves are required in certain positions because I learned how the opening worked

(Edit: Even now after I've learned 15+ moves of Catalan theory, I still haven't seen an opponent stick to the line I knew after move 8, but the positions are similar enough that I know how to navigate them)

Sorry for the lengthy post, just my opinion

In my opinion you shouldnt play too many setup openings because they aren't ambitious: you won't fight for the full point by making the same moves every time and you are likely to miss eventual tactical mistakes by your opponents! I think beginners should know a bit of theory because opening principles don't always work, let me make an example: If a beginner knows principles but no theory at all, he's likely to find himself in this position as Black: 1e4 e5 2 Nf3 Nc6 3 Bc4 Nf6 4 Ng5 d5 5 exd5 Nxd5? Nxd5 is a natural move, yet it loses to 6 d4! Another example is 1 e4 e5 2 Nf3 Nc6 3 Nc3 Nf6 4 Bc4? This natural move is considered bad because it lets Black equalize after 4 Nxe4! I have seen many beginners limiting theirselves by following opening principles to the letter and by playing setup openings without noticing the opponent's moves

Youre comparing GM level play against what beginners should know. 

I agree.

I just like to add that people learn differently. In my case, theory works best for me because of my engineering background. Chess keeps me interested in that I can critically think, analyze and apply “first principles” to me is “for dumb people” as quoted in my industry. 

I see math , algebra, decimals, equations, permutations , combinations and c++ programming lol.

Its not always the case when it comes to an inclusive chess community. So I respect that everyone learns differently and gets to their goals differently as well.

My point is effective communication is based knowing your audience. When I took up literature, yeah I picked up a “Coles Notes for Dummies”.

at beginner level, you want your student to survive and get exposed to a whole lot of positions.  Theory is almost completely meaninglless because they woudnt be able to punish its deviations.

If system play gets them there, all the better. IF they want to start with 1.b4 and do their homework, thats fine. At this level, just getting them to play a lot and start developing pattern recognition is what matters.making them learn 12 moves of ruy lopez is meaningless. Let them reach at least 1400 before they get into that.

my only objection agaisnt beginners playing systems, is that they wont be exposed to as many different types of positions,  but on the other hand, they may not get blown off the board as often, so you wont need to babysit morale as often.

What's the exact opening/variation of this called? I remember seeing it once but I've forgotten. Seems a bit intriguing to me

I completely agree with your point. What openings would you recommend to me perhaps (800-900 range)

#31
Recommended to defend 1 e4 e5 and 1 d4 d5 as black and to open 1 e4 as white.