?! Black should not let white play cxd5 when the knight is on d7. This is known to be strong for white. Better is 4...Be7 5 Bf4 where black will fianchetto the LS Bishop and have a fine game.

It's playable. Black does well after playing c6 and can get in an early e5 where the c-pawn will hang.

There's an interesting endgame variation where black sacs the c-pawn. It's decent for black. 1.d4 d5 2.c4 e6 3.Nf3 Nf6 4.Nc3 Nbd7 5.Bg5 Be7 6.c5 O-O 7.e3 Ne4 8.Bxe7 Qxe7 9.Qc2 Nxc3 10.Qxc3 b6 11.cxb6 axb6 12.Qxc7 Qb4+ 13.Qc3 Qxc3+ 14.bxc3 Ra3

?! It's one way to play, but the weak dark squares do make white's life harder.

?! This is rather serious. Better to let white make that exchange. Also, black needs to develop and consider exchanging white's good knight.

?! Black's pieces are at home.

? Nxg3 is an improvement. Both sides have equal chances, though black might not want to be so materialistic as all this. Notice that Qf2 doesn't work on account of fx3 in this variation.

White is better.

?! Qh3 covers the g2 square and thereby protects the critical h-file.

!! Black plans to give back the exchange. It will hopefully prove too tempting for black to resist.