I missed this thread the first time around, due to vacation...
Palliser (2007)
Fighting the Anti-Sicilians has good coverage of the ...e7-e6 systems mentioned here, in his Grand Prix Attack chapter.
1.e4 c5
A) 2.Nc3
A1) 2...Nc6 3.f4 e6 4.Nf3 d5 5.Bb5 Nge7 with a big branch here.
A2) 2...e6 3.f4 d5 4.Nf3 dxe4 5.Nxe4 Nc6 6.Bb5 Bd7
B) 2.f4
B1) 2...d5 3.exd5 Nf6! 4.Bb5+ Nbd7!? (his marks)
B2) 2...e6 3.Nf3 d5 4.Bb5+ Bd7
One thing I was surprised to learn is in line A1 black is doing well after 6.exd5 Nxd5 7.Ne5 Bd7 8.Bxc6 Bxc6 9.Nxc6 bxc6. (A game he quotes is A.Kosten - S.Cvetkovic, Belgrade 1988.
) Palliser thinks black is fine everywhere, but it might be useful for English Opening players to have a look and see if they could be more than fine with reversed colors and an extra tempo.
RE 1.e4 c5 2.Nc3 Nc6 2.g3 g6 4.Bg2 Bg7 5.d3 e6 6.f4 Nge7 7.Nf3 O-O 8.O-O d5
I consider this line to be ideal for black, and always avoided it with Smyslov's 6.Be3.
RE 1.c4 e5 2.Nc3 Nc6 3.e3 f5 4.d4 Nf6 5.Nge2 d6 6.g3 g6 7.Bg2 Bg7 8.O-O O-O 9.Nd5!?
The real question here is, why did white castle? Wouldn't it be better to play 8.b3 O-O 9.Bb2 (maybe even 9.Ba3!?) ... ? White can develop flexibly on the queenside while if black wants to get anything going on the kingside he needs to make a commitment. See e.g. Anand - Kasparov, New York rapid 1994 (
http://www.chessgames.com/perl/chessgame?gid=1018557). Kasparov said he was doing well until the mistake 16...d4. It's all very reminiscent of the ...e6/...g6 system against the King's Indian Attack.