I always get these wrong, but I'm going to say it's bad, or at least not particularly good; the move brings a white pawn usefully toward the centre, gives up the two bishops, without even requiring a tempo from white with a3 (which may become available to a white bishop). And it initially seemed to me that black would be better off playing 7...dxc4, intending to use d5 as an outpost for his pieces. However, this is quite hard to organize since a future ..Ne7 will imply that a3 must be met by ..Bxc3+, a future Qb3 may oblige the same, and the c3 and f4 knight cover d5 nicely, as might white's LSB shortly. So although I still think ..Bxc3+ is not best, I'm not sure what black's best alternative is.
I'm with Tom.I think the b file will give white play on the queen side, and it isn't clear black can exploit white's central pawn structure.The Bishop, whilst unprotected, doesn't seem tactically weak. And the same move (Bxc3) will probably be available to black next move with the exception of Qb3 or Bd2 but a queen or Bishop on c3 doesn't seem any better for white.Ne7 a3 Bxc3 seems a plausible line, as black I'd be much happier with a pawn in a3 since it will take an extra move at least to get the Bishop onto the a3-f8 diagonal and then presumably to d6.I'm thinking Ne7 Bd3 is more pertinent, white doesn't need to force the issue on c3 with a3 since castling will break the pin before it becomes an issue. But in this position Bxc3 is still available.But I don't play closed games because I don't understand these things either.