From White's perspective I like 1.e4 c5 2.Nc3 d6 3.f4 g6 4.Nf3 Bg7 5.Bc4 Nc6 6.o-o with an early f4-f5 best:

a1) 6...e6 7.f5 gxf5 8.exf5 d5 9.Bb5 is interesting, while 8.d3 Nge7 9.Ng5 (9.Bg5 f6 10.Bd2 o-o or 9.Qe1 o-o 10.Bg5 f6 and 11...Ng6) h6 10.Nxf7 is dangerous, but dubious.

a2) 6...e6 7.f5 exf5 8.d3 Nge7 9.Qe1 h6 10.Bd2 Be6 (Ne5 11.Nxe5 unclear) 11.exf5 (11.Nd5!?) Bxf5 12.Nh4 and 13.Nd5. White has won a tempo compared to Glek-Gavrikov, Minsk 1983.

b1) 6...Nf6 7.d3 e6 8.f5 gxf5 (exf5 9.Qe1 o-o 10.Qh4 fxe4 11.Ng5 or 11.Bg5, but not 11.dxe4 Be6!?; o-o 9.fxg6 fxg6 10.Qe1 d5 11.Bb3 is unclear) 9.Qe1 o-o 10.exf5 unclear.

When I still thought the GPA a viable option (thanks to Ostapbender I don't anymore

), I found it harder to prove compensation after ...gxf5 than after ...exf5. Here I seem to disagree with Lane.

b2) 6...Nf6 7.d3 o-o 8.f5 Nd4 (gxf5 9.Qe1 e6 and e6 9.fxg5 transpose to b1) 9.Nxd4 cxd4 10.Ne2!? with interesting play.

In the famous games Short-Gelfand, Brussel 1991 and Anand-Gelfand, Wijk aan Zee 1996, Black was not necessarily worse after the opening. Moreover I suspect, that 8...o-o 9.f5 gxf5 (again!) is playable for Black. With a knight on g6 (even if the ugly ...f6 has to be played) I don't see, how White gets an attack.

Neither do I like 6...Nf6 7.d3 o-o 8.Qe1 Nd4 (e6 also is good) 9.Bb3 Be6 for White.

There is also 5.Bb5 (+) of course. Generally speaking I must say, that I find Lane's book not very helpful on several important lines.