Been there, and, man, have I done that!

Too right, Joe! Queen's Indians usually feature the Queen's Bishop fianchetto, instead of the King's.

... and that's fair enough, Joe. The Benoni is sharper and riskier, but that has to be set against what you are comfortable and familiar with. These days I prefer the Benoni myself, never having achieved much of a record with the Nimzo-Indian (though from an aesthetic standpoint, I find the Nimzo - despite its unlovely appellation - a very pretty opening).

I think the choice between the Benoni and Nimzo-Indian is an intuitive one that depends on the type of player you are. The Benoni leads to black trying to undermine the white center early on and some sharp lines often ensuing while the Nimzo is more positional and long-term, trying to create a pawn weakness to capitalize on in the endgame; thus being more quiet and leading to some interesting theoretical battles (like Nimzowitch vs Judowitch/Gligoric/Tal).

Qc2 is also very popular and it has my preference over e3, though I sometimes play e3 in a latter stage. Keeps the option for an immediate e4 later on open.

I come down in Joe's favour, here. Against players of about or slightly stronger level, staying with what you know will develop further familiarity with and ability to handle the types of games resulting. You really have to build a solid foundation for your game, and at early development levels this implies (I believe) a small main opening repertoire, with a few change-up lines for the sake of variety.

I don't like that reasoning. If you do that, how will you improve? Sure, trying something new might get you destroyed the first couple of times you do that, but if you can figure out why that happened, you can prevent it next time and hopefully even learn new things about chess in general. EDIT (because I can't seem to simply respond): good point archduke. I didn't fully consider jkarps level of play in this. I guess what threw me off was that he is very familiar with these type of openings, given that I myself started with very standard "golden rules" opening play before getting into more dynamic asymmetric themes. Of course this is not at all the path that everyone takes, so I should apologize for being too presumtuous

A small personal anecdote to prove Martenp: I was shown the 3 golden rules,etc, when I was younger, but I simply lost interest in the game when everything started with the italian opening, the Ruy Lopez, though not very exciting, being a small variation. But in a small blitz tournament with other kids then my age I tried the King's Gambit, and was crushed several times when I thought I was kasparov and made very (unlike him) questionable piece sacrifices. But over the course of the tournament, my naive optimism transformed to realism and true, reasonable, but aggressive play that in fact not only won the tournament but also taught me that opening diversity does not necessarily mean hours of preparation but rather feeling positions that your intuition tells you is YOURS, where you are strong wherever your opponent is weak, where you are confident and aggressive, where he is timid and defensive. And every defense has a weakness...

I thought I'd respond to martenp here. The point of a smallish repertoire early on in one's development is to improve within those limits. Look, you can take on far too much in attempting to develop your game, somehow you have to set yourself limits and attempt to improve within those limits. Later on you can broaden your scope. The idea is to develop a familiarity with the types of game arising from the openings you play. That familiarity implies an improvement, right there. But you are also trying to develop your general tactical and positional play. It is in my view too distracting continually to attempt unfamiliar openings whilst trying to develop your general play and to consolidate what you have learned. Bear in mind, I'm talking sub-1500 level players, here. Once you have improved past that level, then you might well be advised to broaden your repertoire, but even then I would recommend you don't go nuts about it.

c5 seems natural, and Ne2 prevents the doubled pawns. But all this is overstudied by theory (which I don't see!).

The virtue of a3 here is that it resolves the tension about c3, if White isn't comfortable with it.

Ne5 is another idea. EDIT: ...defending c4, so that white doesn't need to open the way for Qxc3. Qb3 is also possible.

By theory, game's databases, etc.

In this case, via the opening databases on Gameknot

Interesting battle, white has the central majority, but black has a nice space advantage on the queenside, if white manages to push e4 I like his chances, but it takes some time.

I think this is rather a risky line for Black, strategically speaking. His Q-side majority might not find it easy to obtain much leverage. Meanwhile, the d5-pawn is a bit vulnerable, especially to a well-timed e4-break. An alternative, 10...cxd4, would have left Black with an isolated d-pawn and a baddish bishop, so I can see why that wasn't an attractive option. Maybe something else had to be sought?