The idea of 10...b5 versus 10...b6 is not to give the chance to white's knight to grab a pawn before his own unavoidable capture.
Additionally it gives black the ...b4 resource in certain variations.. Of course the move has its drawbacks too.. For instance the Ba6 move after white's 00 will not be played with a gain of tempo anymore
11. a3 Bb7, 12.d5 Qxa8 (better than Bxa8 since the bishop had to be redeployed to b7, while d8 is not a good square for the queen), 13. Qe2 e6, 14. 00 is unclear
11. Nxb5? Qa5+, 12. Nc3 Nxe4, 13. 00 Nxc3, 14.bc3 Bb7 and black is better..
11. 00 Bb7, 12. Nxb5 Bxa8, 13. f3 d5 and black is more than OK
11. Qe2(d3) a6
11. f3 Ba6!, 12.de5 Nxe5, 13.Nd5 Nxd5, 14. Qxd5 Bh4+, 15. Kd1 Bb7, 16. Qxb5 Bc6, 17. Qb3 d5! /+ Canneva  Leignel 2003
11. dxe5 Nxe5, 12. 00 Bb7, 13. Nxb5 Bxa8, 14. f3 d5 is Huber  Schaeffer 1997 and must be white's best try.. He has 3 pawns and a rook for two pieces but black's pieces are much better positioned.. Therefore, unclear..
The above is based on Bauer's book..
He mentions some other choices for white's 12th move in the last line, but none changes the general evaluation.. At White's best, the position is unclear and maybe a little bit easier to play with black..
Personally I have scored 100% with Black after 10...b5  3 out of 3, two of those games were over by move 20 (FIDE 1800ish opposition)
