Move 5... Nxb5 is a known mistake in theory forever. The only playable moves at this point are Qe7 and Bc5. Most GMs seem to think that after such moves there is no harmful lines for black and perhaps white might be in trouble. I believe Jan Pinski in his book gave Qe7!! saying white might have quite a load to deal with. Indeed, I agree with short that L'Ami shouldn't play an opening he knows nothing about!

At Corus Short scored a quick win, here is the game:

(17) Short,N (2663) - L'Ami,E (2603) [C48]
Corus B Wijk aan Zee NED (3), 19.01.2009
1.e4 e5 2.Sf3 Sf6 3.Sc3 Sc6 4.Lb5 Sd4 5.Sxe5 Sxb5 6.Sxb5 c6 7.Sc3 De7 8.Sf3 Sxe4 9.0-0 Sxc3 10.dxc3 d5 11.Lg5 Dd6 12.Te1+ Le6 13.Sd4 c5 14.Sxe6 fxe6 15.Dh5+ g6 16.Dg4 Kf7 17.c4 d4 18.Df3+ Kg8 19.Lf6 1-0

Position after 5.Nxe5:


I would like to see Mikhalevski take the chance to fill a gap in the section´s opening book, because already 5.Nxe5!? isn´t mentioned there. The move isn´t new, there are several between Spielmann and Rubinstein in the 1920s. Since then 5...Qe7 is preferred by theory, and I wonder why L´Ami played 5...Nxb5. The following moves are all natural, but after 9.00 I don´t like the black position, though black was by no means lost by then and didn´t have to go down in 19 moves.

Most interesting would be how Short would have continued after 5...Qxe7 6.f4 Nxb5 7.Nxb5 d6 8.Nf3 Qxe4+. According to Pinski this is almost winning for black following old analysis by Keres. Yet Short played this line already in 2001 against Khalifman, who deviated with 8...c6?!, so he must have something up his sleeves. Any ideas?