Interesting stuff. What's your preferred approach after 3 Nf3 Bb4 4 Nc3? I wouldn't myself want to go 4 ...Nf6 unless I had to, but that's no doubt just personal preference (4 ...d6!?; 4 ...Bc3?!).

Yes, that's one of the main points. White is better off without his Knight on -f3 in these positions. 
In the Keres Defense, 1. d4 e6 ; 2. c4 Bb4+ ; 3. Bd2 a5 ; 4. e4 and 4. Ktc3 are rather sterner tests and if I recall correctly, ...Ktc6 is on the contrary usually played early on in the slower lines following Ktf3.

Presumably the only 'point' could have been to get White to commit himself, but I guess just 4 Nc3 shows it to be fanciful rubbish ...

Hi Uberdeker and good to have you back! I have to confess I never got round to taking a proper look at your eponymous defence but I'm enjoying getting to grips with it now!

One thing I might mention is that while you've been 'away', Stefan Buecker, linksspringer and I have between us restored the Bogolyubov Defence 1 d4 Nc6 to Full Capstan Strength (see http://www.chesspub.com/cgi-bin/yabb2/YaBB.pl?num=1271505369/0)! Not that this makes 1 d4 e6 2 c4 Nc6!? any less interesting ...

3 Nf3 is, I guess, a good try to spoil Black's fun, whatever its objective status. Messing around with Houdini, I noticed that one of the bigger Arena opening books actually produces the weird 3 ...f6!?, presumably with the idea ...Bb4 and ...Nge7 (or ...d5/...Bd6/...Nge7). Of course I make no claims for this, but it seems to be in the spirit of the opening and might not be so daft? ...

Nice thread! This is just the sort of line that could make it into Secrets of Opening Surprises.

I don't believe in 3.a3, to be honest; after 3...f5 4.b4 g6 5.Bb2 Bg7 6.Nf3 Nf6 Black will play ...0-0 and ...d6 with a good version of the 'Christmas Tree' line of the Leningrad Dutch. 

Against 3.Nf3, 3...Bb4 is interesting, since after 4.Bd2 Qe7 Black is not worse off than in a normal Bogo-Indian (where ...Nc6 is usually played). If 5.e4, then 5...d5 or 5...Nf6 is playable, though I don't know if objectively Black achieves full equality. Best to stick with 3...Nf6 I suspect.

White doesn't have anything after 3.d5 Ne5 4.e4 e6, as demonstrated in the game Solomon-Teichmann. 

After 3.Nc3 Bb4 it's hard for White to avoid a transposition to a quirky but playable Nimzo or Dutch. With 4.e4 not cutting the mustard, 4.d5 is the other independent move, but after 4...Ne5 (remarkably this obvious move is a novelty) 5.Nf3 (5.e4 Nf6 transposes to a promising line of the Tango) 5...Bxc3 6.bxc3 Nxf3 I think Black equalises regardless of how White recaptures. Against 4.e3 I would be inclined to play 4...f5, but against 4.g3 I would prefer 4...Nf6 5.Bg2 d5. 

3.g3 is an interesting move since a kingside fianchetto is useful against both the Dutch and the Tango. Then Black can either transpose to the Catalan with 3...Nf6 4.Bg2 d5 5.Nf3 dc4 or a Bogo-Indian with 3...Bb4 4.Bd2 Qe7 5.Bg2 Nf6. 5...f5 may be playable but I don't trust Dutch setups with a knight on c6 when White has fianchettoed his king's bishop. True, 1.d4 f5 2.g3 Nf6 3.Bg2 e6 4.c4 Be7 5.Nf3 0-0 6.0-0 d6 7.Nc3 Nc6 was covered in 'Dangerous Weapons: The Dutch', but if White is well prepared then he has the advantage.

Funny that you mentionned 1. ...Ktc6 , 1. ...f5 and 1. ...e6 ; 2. c4 Bb4+, as there are elements of all of these in the UD. Add to this the "King's Indian centre" -d6/-e5(/-f5) that crops up in many cases, coupled with your disbelief in White's chances in the critical lines, and it would that appear that the UD is quite your cup of tea. Are you planning to wheel it out some time?

Fair enough, Frendo. Out of curiosity, I'd like to know what you play with Black vs 1.d4 and if you play 1.d4 yourself, do you prefer to face the Nimzo or Queen's Indian, and if the former, which lines to you employ ?

[quote author=Frendo link=1142445517/30#35 date=1167166557] Just because another defence allows good variations that your defence avoids doesn't mean that your defence doesn't allow White other good options.  [/quote]

 Well, this is what I was getting at. The most critical lines must be examined first. For 1. d4 Ktf6 ; 2. c4 e6, this is 3. Ktc3 rather than 3. Ktf3. For the Two Knights' and Keres Defences these are the lines I listed in my last post (there may be a few more). In the UD, the most obvious are, in order of importance, 3. Ktc3 Bb4 ; 4. e4 , 3. Ktc3 Bb4 ; 4. d5 , 3. d5 and 3. e4. 
Only when these have conclusively been shown not to pose great problems for Black, can we say that 3. Ktf3 is the best attempt at an edge. One example of such a process is the evolution of the Najdorf. The critical 6. Bg5 has been almost abandonned by attacking players in favour of 6. Be3, which leads to a slightly less effective version of the Keres Attack (6. g4) vs the Schevenigen. I believe the same to be true for 6. Bg5 against the Underground System (1. e4 c5 ; 2. Ktf3 Qc7 ; 3; d4 cd ; 4. Ktxd4 a6 ; 5. Ktc3 Ktf6) : the critical test of Black's move-order, but not the best/main line. But only through detailed analysis can one arrive at such conclusions. 
As you have said yourself, the jury is still out ! But few of my opponents have so far been willing to put the soundness of the concept to the test.