yes yes, I saw that, very nice! in that variation white can, of course, trouble black a bit more with 15.Bxa6(instead of 15.Nxc8) ...bxa6 16.Rc1, but after ...Qb5(16...Qb6 17.Nd5 Qxb2 18.Re2 is worse) black still has a clear edge
I don't think 10.c5 is very frigthening any more!

What about the position after 20...Bf8 21.b4 a6(I don't like ...Qa4 22.a3 Nc6) 22.a4 Rc8 23.Qd2!?(23.Qxc8!? is very interesting but black should be OK if he knows what to do) ...Rd7 24.Qb2 I have looked at this for a while, but I am not sure what is the best way forward. Black should probably choose between ...Nc6 and ...Nf7. White has a space advantage, but black seems no worse... What do you guys think? Are there any improvements earlier in this line or should black stick to this variation?

(Bear in mind that we are already 13 moves into a complicated variation!)

come on, smyslov fan....You should have a position in mind if you are claiming that white is better.

Sounds good, 

I'll send you one in the near future!

Teyko, 

You're right, 16...f6 coupled with ...Qf7 just wins for Black.

I've found a line (14...Qxc5 15.Bxa6 bxa6 16.Rc1 Qb6 17.Nd5!?) in which White is barely hanging on, but why bother?  White sacked the pawn with winning in mind, not with the idea of scraping hard for a draw two pawns down.

There are almost certainly improvements for White in the 10.c5?! line, but White has plenty of game in 10.dxe5 variation.

WOW!   

It's fantastic to see such concrete analysis in this forum!  I know I'm jumping in rather late, but here goes....

I wonder first at White's move order 7.Be3 (Gligoric) is less flexible in my opinion than castling immediately.  If White plays 7.0-0, he invites 7...Nc6, and I like the thought that White has at least three good moves at his disposal.

Secondly, I am very suspicious of 10.c5 when 10.d5 or 10.de5! still looks fine for White.

However, looking at the analysis, it seems that White doesn't take advantage of the possibilities of trading off the dangerous light squared bishop in various lines.  I like the latest line that teyko has given for Black, but I wonder if White can still gambit the c-pawn since that seems to be the whole point of the variation.

Here's my idea (I haven't checked it with Fritz yet, so it's probably got major problems):

12....Kh8 looks ingenious if Black can afford the time, but after 13.Re1 Qf5 14. Ne7!? (instead of 14cd6?!) Qc5 15.Nc8 Qc4 16.Ne7 followed by 17.h3 or Rc1 White may just have enough compensation for the material lost because Black's remaining minors are out of action.  Yes, I know that the Ne7 has nowhere to go at the moment, but it's a bloody big nuisance there and may require an exchange sac to get rid of it.

Anyway, I'd like to do some serious analysis of 10.de5 de5 which I always thought was the critical test of 9.Qe8 when you're ready to move on from 10.c5?!