I'm surprised this guy is rated so high making such crap moves as these. The annotator is correct. Black will be able to develop with tempo by generating threats against White's queen. Also, it brings the queen to an inferior square, where it will have to move from anyway, and cramps his own game by denying his knight a strong square, and as a result hemming in his own bishop, at least temporarily.

I still think that a move would be better spent developing rather than only taking away one square from one piece, which isn't all that good to begin with. The main purpose of moving the bishop to g5 would be to pin the knight, which it won't after black goes Be7. The only other purpose would be to possibly take the f6 knight for the bishop, which isn't recommended since it's the c6 knight that supports the center, not the one on f6, and then black would have the two bishops. The bishop left on c1 can still be useful along the c1-h6 diagonal, and the b1-knight jumping to d2 and then to either b3 or c4, is not going to hinder the c1-bishop long enough to make a difference.

I don't see why this was done. The knight didn't stand very well where it was anyway. IMO, black should have just castled here. Black's position is enviable, and white has no immediate threats. Continuing his development and bringing his king to safety should have come first. Black could then have launched an attack once his spacial advantage was consolidated. Now the c6 pawn is blockaded and backward, and a white rook on the c-file is ever more dangerous.

The trappy 19...Ra4 with the idea 20.Nxa4 Qxb4 looks good, but it's hard to see that it's really better than the immediate capture on a2.

Black would have had immediate threats against g2 and c2 as well. E.g. 25...Raa2 26.g4 (say) Rfc2 etc.