@OldGrizzly: The line 11...Nd5 12.Bc1 was said to be the main line, and 11...Nd5 seemed plausible at first. I had hoped that after 11...0-0 White might get new ideas (better than 11...Nd5 12.Bc1, when the retreat of the bishop costs time). Something like 11...0-0 12.Qd2 Nd5 13.Bg3, perhaps, or 12.Nf3 Nd5 13.Bg5. But I didn't check anything, and now your proposal 11...0-0! looks just strong.
SWJediknight wrote on 03/02/09 at 20:01:44:
How about the immediate sacrifice on e6? I haven't seen any references to it in that article or in this thread so far. (8.Ng5 Bg6) 9.Bxe6 fxe6 10.Nxe6 leaves Black a piece up for a pawn, but Black will be forced to forgo castling rights. After 10...Qd7 [...] 11.Bg5 Kf7 (11...Qxe6 12.Re1 Ne4 13.Qf3 Nxg5 14.Qg4 leaves Black with a rook and two pieces for the queen, but White maintains an attack; 13...Kd7 will leave Black's king exposed on the queenside) 12.d5 Be7 13.Qf3 is likely to net a second pawn for the piece, with unclear attacking chances. Fritz doesn't think White has full compensation, but Black's position will be difficult to play at least.
An inspiring idea! I've had (8.Ng5 Bg6) 9.Bxe6 fxe6 10.Nxe6 Qd7 11.Bg5 Qxe6 12.Re1 Ne4 on the board, but didn't really believe in it. However, you are right that the attack deserves attention. For example: 13.Qf3 Nxg5 14.Qg4 Kd7 15.Rxe6 Nxe6 16.Re1 Bf7 17.Qf5 Bg8
18.Kh1!! (stronger than the immediate 18.d5, which later allows the freeing Bc5+) 18...Na6 (another weird line: 18...g6 19.Qf3! followed by d5, =) 19.d5 cxd5 20.Qxd5+ Kc8 21.Rxe6 Nc7 22.Re8+ Nxe8 23.Qf5+ Kc7 24.Qxf8 Rd8 25.Nb5+ Kb8 26.h3 Rc8 27.Qf2 b6 28.Qf4+ Nc7 (28...Kb7 29.Qe4+) 29.Nxc7 Rxc7 30.Qf8+ Kb7 (30...Rc8 31.Qf4+ =) 31.Qf3+ Rc6 32.Qg4 Rc7 33.Qe4+ etc. draw.
Many side-lines remain to be checked, e. g. 13...Qf7 (maybe better than 13...Kd7). In any case the variation seems to be a lot of fun.