In my analysis file on the Fried Liver I gave the following:
(9.0-0 c6 10.d4) Kd7 "may be slightly less dicey
(than 10...Kd6), but after 11.a3 (White should not go in for 11.Bxd5 Nxd5 12.Nxd5 cxd5 13.Qxd5+ Kc7 14.Qxe5+ Bd6, as although White gets a third pawn for the piece, it is clear that Black is over the worst) and now 11…Nxc2 12.Qf5+ Kc7 13.Qxc2 or 11…exd4 12.axb4 dxc3 13.Bxd5."
In the line 11.a3, I obviously missed 11...Nxc3, but following one of Sloughter's lines: 12.axb4 Nd5 13.Bxd5 cxd5 14.Qxd5+ Kc7 15.Qxe5+ Bd6 (15...Qd6 is met by 16.Qe8) 16.Qxg7+ Bd7 17.d5 b6 18.g3:
I slightly prefer White here due to having four pawns for the piece, although White has to be careful of Black's attacking chances on the kingside.
I agree that 13.dxe5 Kc7 14.b5 is quite messy and roughly equal, e.g. 14...Be6 15.bxc6 bxc6 16.Rd1.
In over-the-board play, though, I would rather be White.
I agree that the generally-neglected 9.Bb3 is also promising for White. A couple of years ago I regarded 6.Nxf7 as inferior to Lolli's 6.d4, but today I'm not so sure.