That is an interesting idea which I have looked at briefly. You have to be somewhat careful with move orders if white for example plays 0-0 instead of c3. Ba6 is really not effective until c3 has been played since white otherwise has the option of transposing to the Queen's Indian with c4. In those positions, I prefer to have my bishop on b7.

I don't think you should be so quick to dismiss 3...d5 as boring.  I have noticed that many professionals have a preference for setups with ...d5 against the Colle, and it's surprising how easily white can be outplayed if he makes routine moves.  I made a few remarks on some games in the "COLLE-ZUKERTORT" thread.  Sometimes when playing something offbeat to imbalance the position, you actually give white a chance to obtain an opening advantage in these quiet queen pawn openings.  In a way, I think this justifies white's opening choice.  There's something to be said for playing what you think is objectively the best move in a position, even though it may initially a have symmetrical character.

At first, the possibility of having a boring symmetrical structure like this against a weaker opponent bothered me.  You can see my thread on "The Benoni Avoided" in the Nimzo-Benoni section.  In retrospect, many of my comments and assessments in that thread are naive, though not completely unjustified.  Looking back, it shows how one can get a warped picture, when making generalizations after encountering difficulty in a specific area.  In forums like this, I have seem extreme cases of this nearsightedness from "analysts" who specialize in dubious gambits or development schemes.  Often in doing so, these "analysts" consistently misevaluate positions in order to justify some fixed belief; and as a result, hinder their own progess as their beliefs prevent them from developing a universality of style.  It is pitiful to see someone spend hours analyzing a hopeless position devoid of activity resulting from a gambit, when they would achieve more active positions from the Colle.

OK, that perhaps was a bit of a tangent, but the point I am trying to make is that if you consistently try to avoid certain types of positions on the notion of a belief, you may be hindering you own development.  

When I was stuyding these Colle positions, it was interesting to find how easily natural moves on white's part could easily backfire when black retained flexibility.  Often in these systems, players with white have a one-sided plan, usually an attack on the king.  If you have studied these positions carefully, you can use this knowledge to great advantage against weaker opponents.  For in these ...d5 systems, there is a psychological trap.  Often when white sees ...d5, he is happy to play his programmed attack (or rather the attack he "believes" in, to the use the terminology of the last paragraph).  When his attack backfires, due to his lack of awareness of the subtleties in the position, he becomes worse without realizing why.  At such a point, it is all too common that he will blunder due to a gross misevaluation.  This is a very efficient way to play against weaker opposition!

The restrained Queen's Indian development is also a good way to play.  Also, the Reti development is sometimes effective, though I would not play ...g6 until I have played ...b6 first.  These are two games by Mihail Marin that illustrate a flexibility of approach.    In the first game, he plays ...d5 against a Koltanowski stonewall setup.  In the second game, Marin plays Reti-style against a Zukertort-type setup.  Often, a good way to find an antidote to a theory-avoiding continuation is to do a search for a strong player whose style you identify with.  Usually you can find a good way to play for a win against weaker opposition.  

I would highly recommend the book The Colle-Koltanowski System by Valerij Bronznik, which is a well researched book with excellent explanations.  The book is subtitled "Deceptive Peace behind the Stonework."  At first this might sound laughable, but I really think there is some truth to this statement in that there is more to these positions than appearance suggests.

[Event "Manresa op"]
[Site "Manresa"]
[Date "1995.??.??"]
[Round "6"]
[White "Fernandez Aguado,Juan Luis"]
[Black "Marin,Mihail"]
[Result "0-1"]
[Eco "D05"]
1.d4 Nf6 2.Nf3 e6 3.e3 b6 4.Bd3 Bb7 5.0-0 Be7 6.Nbd2 c5 7.c3 d5 8.Ne5 0-0 
9.f4 Ba6 10.Qe2 Bxd3 11.Qxd3 Qc8 12.g4 Nc6 13.g5 Ne8 14.Qb5 Nxe5 15.dxe5 c4 16.Nf3 g6 
17.Nd4 Ng7 18.Bd2 Bc5 19.Kf2 Qd8 20.Ne2 h6 21.h4 Nf5 22.Rh1 a6 23.Qa4 b5 24.Qc2 h5 
25.Ng3 Nxg3 26.Kxg3 Qb6 27.Kf3 Rfb8 28.a3 a5 29.Rhb1 Kg7 30.Ke2 b4 31.axb4 axb4 32.Rxa8 Rxa8 
33.cxb4 Bxb4 34.Bc3 Ra2 35.Bd4 Bc5 36.Bxc5 Qxc5 37.Qc3 Qb5 38.Kf3 Qb7 39.Kg3 Qb6 40.Kf3 Ra4 
41.b4 Qb5 42.Qb2 Qb7 43.Kg3 Qa7 44.Qc1 Ra3 45.Kf2 d4 46.Qxc4 dxe3+ 47.Kg1 e2+ 48.Qc5 Ra1 
0-1

[Event "Bad Woerishofen op 18th"]
[Site "Bad Woerishofen"]
[Date "2002.03.??"]
[Round "6"]
[White "Popitz,Andreas"]
[Black "Marin,Mihail"]
[Result "0-1"]
[Eco "A30"]
1.d4 Nf6 2.Nf3 e6 3.e3 b6 4.Be2 Bb7 5.0-0 g6 6.b3 Bg7 7.Bb2 0-0 8.c4 c5 
9.Nbd2 Qe7 10.Qc2 Nc6 11.Rad1 d6 12.Qb1 Rad8 13.Qa1 Ne8 14.dxc5 bxc5 15.a3 a5 16.Bxg7 Nxg7 
17.Qc3 f5 18.Ne1 f4 19.Bf3 fxe3 20.fxe3 Nf5 21.g3 d5 22.cxd5 exd5 23.Rc1 Nxe3 24.Rf2 c4 
25.bxc4 d4 26.Qb3 Ba8 27.Nd3 Rb8 28.Qa4 Ne5 29.Nxe5 Qxe5 30.c5 Bxf3 31.Nxf3 Qxc5 32.Re1 d3 
33.Qd4 Qxd4 34.Nxd4 Rxf2 35.Kxf2 d2 0-1

Excellent replies, thanks! I think I'll try the 2... b6 setup and if 3.Bg5 maybe 3...Ne4. 2...d6 with the idea of 3...Bf5 is also interesting. I'll have a look at them and make my choice!

Thanks again,

Joris

Alumbrado's idea of 1.d4 Nf6 2.Nf3 e6 3.e3 g6 is quite interesting, as Black indeed does not lose a tempo with a later Qe7 and e5 compared to Qe8, e7-e5 and Qe7. I remember having seen Jeroen Piket executing this plan.
Against 1.d4 Nf6 2.Nf3 e6 3.e3 b6 White has nothing better than transposing to a Queen's Indian with a later c2-c4 and Nc3.

I once looked for the Colle to be part of my White repertoire. But when I found out that Black with a setup of g6+Bg7+d6+e5 (pawn confrontation instead of a closed pawn centre) scores over 65 % (!) of games in the Megabase, I did finish my Colle-studies.
But I have no idea how to include this setup in the repertoire of someone who already plays for example in move nr. 2 e6???