I like the Nc3 french, but I'm playing a system based on 1. d4 d5 2. Nd2, which combines the Breyer slav + Tarrasch french + Colle Zukertort. Which... seems to be virtually unknown, however objectively no different than the Jobava but it's surprisingly one of the sharpest / most tactical d4 repertoires I've seen. But this tarrasch french will be the most theoretical line in the repertoire. So far I'm liking the lines here in the Tarrasch quite alot, though. I think that Chistayakov line is going to crush and the f4 line... I'm happy with that too.
Thanks for the tip on the open tarrasch - I haven't made up my mind there yet but will take a careful look at your suggestion.
What is the point behind 1.d4 d5 2.Nd2? There really isn't one. It isn't like e4 is a threat at getting a big center. Black will just trade on e4.
Now in The Colle - Move by Move, Lakdawala introduces a line he calls "An Ostrich in the Sand" where he demonstrates 1.d4 Nf6 2.Nd2. This is different. Here White does threaten to get the big center with 3.e4, and the difference between this and 2.Nc3 is if 2.Nc3 and Black does play 2...d5, you are stuck in the territory of the Jobava, Veresov, or possibly the Barry Attack. By playing 2.Nd2, if 2...d5, White can go back into the Colle, most likely the Koltanowsky (c3) as the early commitment of the knight could lead to move order issues in the Zukertort.
Of course, if Black brings the Bishop out to f5 or g4, white will need to act fast and play something like c4, as is the case after the normal 1.d4 d5 2.Nf3 Nf6 3.e3 Bf5, where White's only good move is 4.c4, when 4...c6 is the Slow Slav.
Let me start by saying I generally place little faith in abstract theorizing about the opening, especially the kind done in complete confidence but divorced from any tangible experience. In your case - you just finished extoling the virtues of the Levitskys, so your attitude is even more confusing and obscure to me, but I digress.
People often make the same sorts of claims about the Van Geet, they say you're blocking in he C pawn, it's anti positional. In practice it's outscoring 1. e4 and 1. d4.
While abstractly theorizing I think you can fool yourself into believing you've arrived at the best possible conclusion, when in reality you're lost at sea down in the depths clinging tightly to some seaweed.
Instead of doing that, it's much better to just look at the resulting positions and base conclusions off that.
I don't think Ostrich in the sand is the right name, it's not doing it justice. Rather I call it the Wormsnake. Because you think it's harmless, you think it's a worm, but then it uncoils with c4 or e4 (either is possible, sometimes both) - the position gets chaotic very quickly, it is infact a vicious snake. This is one of the trappiest lines I have seen in 1. d4...
For some comparison, if you've ever played the Breyer slav... it's known for being a very trappy slav sideline. Almost all of the sidelines in the Wormsnake here play very much like the Breyer. The lines which aren't sharp are the known ones - colle, French tarrasch.
Now theoretically... if we wanted to speak in these somewhat-useless abstract terms, we can say Nd2 allows white to push either e4 or c4, sometimes both, while being immune to pins via Bb4, or exchanges that damage the pawn structure. Whites DSB also defends b3, but isn't locked in. Alternatively, White can Fiancetto on the kingside while pushing c4 and not worrying about dxc4. Likewise white can play Qb3, and dxc4 doesn't tempo the queen. The queen here can also often rotate to the kingside, it happens. Or white can play b3 and, and since his knight doesn't block the bishop he will exert good control over e5 / his knight is already rotating over to reinforce Ne5.
So yes, there are ideas in the position. And it's actually an extremely flexible system.
The move is fine against 1... d5.
In the Colle positions Nd2 is facilitating b3 > Bb2.
I'm only transposing into the Colle in 2 lines - when black plays an early e6, or g6.
Against e6 - this is actually the most winning line for white in the Colle. White is up by 10 points here, which usually isn't the case. I think the reason for this is... a) black can't develop his bishop outside the pawn chain, b) he also didn't push c5 early which is one of the more problematic moves for the Colle.
Against g6... Actually, I like the colle against g6.
I've played many of these offbeat zuckertort-type queens pawn systems - London, Jobava, Veresov, Torre... g6 is usually the most annoying line to face in these systems. The game is usually very slow and grindy, and you'd have been better off in a pirc or KID/Grunfeld. Often times your bishop ends up not doing very much, getting chased away by a knight or lasering thin air.
Here in the Colle, with b3 > Bb2, you're directly opposing blacks kingside bishop. It leads to whole-board positions which I prefer over the attempts at pushing queenside pawns while blacks kingside bishop is lasering you the entire time.
So the 2 Colle lines I get from Nd2 are probably the 2 I would prefer, if I had my choice of Colle lines. I'm not such a fan of the Colle in general, but these 2 lines are the exception.
Overall though, the Colle is not a common transposition here. The Breyer slav is more common. Because d5/e6 is a french tarrasch transposition after 3. e4
But what you usually get are the original lines. Wich usually involve either an early Qb3, or c5 > dxc5 gambit lines.
So no, it's a fine system. A fine system that has pretty much never been given any attention. And it happens to be very sharp, which is.... exactly what you want in a rare line.
This early c5 line is one of the best lines in the system for white.
Especially followed up by e5 from black, the position you posted. Thats the most common black response but... It's quite sharp, chaotic and obscure. Any time an opening plays like that it favors the prepared player, which is simply always going to be white in this case. The most played move in your position, after e4, is d5... and it is losing already. c4 is possible too yes, and it's also quite good. But e4 is I think is even more volatile.
All the early c5 lines play like that - very sharp, very odd. And objectively they're all fine for white, the worst you'll get is a position that's about +0.00, but still volatile in a way that swings in whites favor.
On that note... lets get the objective claim out of the way. Leela scores the Van Geet -0.02, and the Levitsky -0.10. It scores this Nd2 line +0.00. There is no line I've seen in this system that either doesn't transpose into something known - Colle, Breyer Slav, French Tarrasch - or isn't totally sharp and obscure in a way that is great for white.
Anyway, I like the Van Geet but the one criticism I have of it is... black can take the game into somewhat-known territory in many lines.... c5 will get you a sort-of sicilian, d5 and most people have encountered the chigorin system at some point. As for the Jobava - I think 10 years ago it was a pioneering opening, it was very original and lethal. Today I don't really believe it is anymore. Everyone knows its name now, most people have some idea of what it is / how to play it. In these queens pawn sidelines, or any line where you give up the advantage, a very major part of your compensation is the surprise factor. If that surprise dies off... your opening is losing its most important aspect.
For this reason I didn't play the Jobava, I played the Veresov. In the Veresov, though, there are a number of lines where it can become a "slog". Meaning... it's just an equal game where the two are slogging it out, white doesn't have something sharp, or unusual, that really justifies his giving up the advantage. Sometimes the Veresov pays off, sometimes it doesn't.
There is no line in this Nd2 system that feels remotely like a "slog". There is no easy formulaic response that black is going to know, either. Except if they happen to be one of 12% of french players and find their way into the tarrasch french... that is the only really "known" line.
The Colle transpositions aren't actually that common, but the people who know their colle lines play the anti-colle or the early c5, both of which this system avoids.