The position after 11.c5 has occured 15 times in practice according to Mega Database 2008, with White scoring 63% (+7, =5, -3). However, White's average rating was 179 points higher than Black's, so I would see this result as being a success for Black statistically.
Black is forced to take on c5 (11...0-0 12.cd6 cd6 13.Be2 +/-), when after 12.Qd5, 12...Nc6 has never been played in practice, but your recommended 12...Ng6 has done very well in practice, leading to the following position:

After 13.Bb5, Black can play 13...c6, 13...Bd7 or even 13...Kf8!?, all of which look fine for Black, although the game with 13...Kf8 looks like the most reliable reference. Here is the game below:
Drazic (2465)-Facchetti (2160), Milan Crespi open 2002:
14.0-0-0 c6 15.Qd8 Qd8 16.Rd8 Ke7 17.Rh8 Nh8 18.Bd3 Ng6 and White's two bishops did not fully suffice for the pawn. White continued with 19.Bg6 hg6 20.Rd1 and White eventually drew the game, although Black retained some winning chances throughout.
However, 12.Qc3 is also possible. After 12...f6 13.Rc1 I think White has enough compensation for the pawn, but not enough to claim any advantage. For example: 13...b6 14.Be5 Qe5 15.Bb5 and instead of the previously played 15...Bd7?!, 15...Ke7! looks equal.
Conclusion: 11.c5 is harmless as long as Black is careful. However, note that 10...0-0?! 11.c5! is very good for White.