Browsing through my notes on the MG I found the following analysis by Ben Hague:
<1.e4 c5 2.d4 cxd4 3.c3 dxc3 4.Nxc3 Nc6 5.Nf3 d6 6.Bc4 a6 7.O-O Nf6 8.Bf4 Bg4 9.Qb3 e6 10.Ng5 b5 11.Nxb5 h6
(11...axb5 12.Bxb5 Rc8 13.Rfc1 +-)
(11…Na5 12.Qa4 axb5 13.Bxb5 Nd7 looks like the main line. I'm not sure what's happening here, white can get another pawn for the piece with 14.Nxf7 Kxf7 15.Bxd7, and there is clearly some compensation for the material, but if it's enough is hard to say. My gut feeling is that while there are good practical chances it may not be quite enough)
12.Bxe6
(12.Nxe6 fxe6 13.Bxe6 axb5 14.Bxg4 Nxg4 15.Qe6+ Qe7 16.Qxg4 unclear)
12…hxg5
( 12...fxe6 13.Nxe6 Bxe6 14.Qxe6+ +- )
( 12...axb5 13.Bxf7+
( 13.Nxf7 Nd4 14.Qe3 Ne2+ 15.Kh1 Nxf4 16.Qxf4 Bxe6 17.Nxd8 Rxd8 +/- )
13...Ke7 14.Bh5 Qd7 15.Qf7+ Kd8 16.Bxg4 Nxg4 17.Ne6+ Kc8 18.Qf5 +/- )
13.Bxf7+ Kd7 The critical position
( 13...Ke7 14.Bxd6+ +- )
14.Bxg5
( 14.Nxd6 gxf4
( 14...Bxd6 15.Bxd6 Kxd6 16.f3 unclear)
15.Nf5 Kc7 16.Rfd1 Qc8 17.Qc4 Bxd1 18.Rxd1 unclear)
14...axb5 15.f3 Nd4 16.Qd3 Be6 17.Bxe6+ Nxe6 18.Qxb5+ unclear
I'm not sure that all this works, but I think that at least it deserves analysis.>
End quote.
Überdeker, I have done my best for you. You have work to do.
@Arkhein:
This variation is pretty good for White: 11.Bf4 Qb8 (b4 12.Nd5) 12.Rd2 (Palkovi suggests 12.Nd5) Be7 (b4 13.Nd5) 13.a3 0-0 14.Rad1 Nc5 15.Bc2 e5 16.Bg5 and I won a friendly corr game MNb-Peeters, 1994. Black should have played Rc8. Moreover there is 13.Nd5 (already getting bored?) b4 (Ne5 14.Nxb5! axb5 15.Nxe5 0-0 Zelic-Armanda, Split 1998, 16.Nxf7!? with interesting complications) 14.Rxd6 Nc5 15.Qc4 bxc3 16.Rxe6 fxe6 17.Bxb8 Rxb8 unclear.
No Open Sicilians with an extra pawn here.