Opening Archaeology
First: shortly after my last post, I dug into my settings and found that I actually am 32-bit...so at least I don't have to deal with OS shuffling. The extra computing power would have been sweet though. I'll be editing the post to fix that.
I'm going to make a new tag, "opening archaeology" for intriguing old opening ideas I come across as I'm studying collections of master games. The following Sicilian idea seems a bit dubious to me, but those that find the 2...e6 Sicilians annoying may want to look at it.
In The Book of the New York International Chess Tournament 1924, Alekhine in the footnotes of a couple games reveals that after 1.e4 c5 2.Nf3 e6 he likes 3.Be2!? 
One of the ideas is that it may be possible to avoid playing Nc3 to protect the e4 pawn. In most open Sicilian lines, ...Nf6 attacking the e4 pawn is played early, which prompts White to play Nc3 and block their c-pawn. The Maroczy Bind (pawns on c4 and e4 in the Sicilian) was considered a Very Good Thing at the time.
In this case, if 3...Nc6 (or another move besides 3...Nf6--3...Nc6 is the most common in this position) 4. d4 cxd4 5. Nxd4 Nf6, now 6. Bf3 protects e4:
In theory, c2-c4 would be in the cards now. However, in Tartakower-Steiner, Saltsjobaden 1948 I see that White never got around to it: 6...Ne5 7. O-O Be7 8. b3 O-O 9. Bb2 (9.c4!?) ...d5 10. Nd2 Nxf3+ 11. Qxf3 Bb4 12. e5 Nd7 13. Qe2 Bxd2 14. Qxd2 Nxe5 15. Nxe6 Nf3+ 16. gxf3 Bxe6 17. Kh1 f6 18. Rg1 Bf5 19. Bd4 b6 20. Rg3 Rc8 21. c3 Qd7 22. Rag1 Bg6 23. Qf4 Rf7 24. Re1 Re8 25. Re3 Rff8 26. Rg1 Qh3 27. Rg3 Qf1+ 28. Rg1 Qh3 29. Rg3 Qf1+ 1/2-1/2
In case of 3... Nf6, Alekhine gave 4. Nc3, with the idea of 5. e5 Nd5 6. Nxd5 exd5 7. d4. I started looking at sample lines, and it doesn't seem that simple, but Fritz is giving White a nice edge. I'll leave it to the reader to explore those variations.
I'm unconvinced by the idea of putting the bishop on f3 behind the e4 pawn, and in a few of the games I looked at this indeed became a problem if the pawn remained there. Also, to me the most testing response to 3.Be2 is 3...Nf6, where White ends up playing 4.Nc3 anyways. However, it seems playable, and I think the odds are your opponent would not have looked at these lines.
