I looked at alternatives to your 7.a4, not really finding anything convincing for White.  7.Bb5+ and the weird computer suggested  7.cxb6 axb4 8.Qd4!? keep the game roughly level and unbalanced.  White could also try the suggested 4.b3, but the reason this isn't suggest is that Black isn't commited to ...e6 with his Bishop locked inside the pawn-chain (of course this is playable for White, but not part of the suggested "Killer" repertoire.  By the way, the anti-colle moves (after 1.d4 d5 2.Nf3 Nf6 3.e3) 3...c6, 3...Bf5, and 3...Bg4 are covered a little on the "jeremysilman.com" site.

Greetings to all "Specialists"!

I've been using Summerscale's "Killer" repertoire with White since the book came out, with good results. Finally I could play White without hours of preparation before every game; a big thanks is due to Aaron ! I have found the so-called "Anti-Colle-systems" to be most difficult to meet, and in particular the lines where black plays ...d5 and ...c5, kind of a reversed Queen's Gambit. I was faced with what I consider the most unpleasant of these systems. with white commited to e3, at Gausdal this year (see Martinsen-Bisby below). Summerscale has the idea of making it a reversed Noteboom (1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 dxc4) which is an interesting, unbalanced opening, but in my game Bisby took advantage of not yet having a knight on c6 to leave me with "only" one passed pawn. As the game went I was very lucky to escape with a draw. In my database Black scores very well with 
6...b6 - it seems that Black gets easy development and a constant threat of ...e5 (and ...e4) that White must be ready for, without White getting the dreaded "Noteboom pawns" as compensation.
While the killer repertoire is fun to play I don't really expect a theoretical advantage. But can anyone show me a way to make it at least interesting against this line? Of course White could simply accept that Black will get the pawn back and play a QGA reversed, but it doesn't really appeal to me.

Help, anyone??

[Event "Troll Masters"]
[Site "Gausdal"]
[Date "2003.01.07"]
[Round "4"]
[White "Martinsen,Stig K"]
[Black "Bisby,Daniel L"]
[Result "1/2"]
[Eco "D04"]
1.d4 d5 2.Nf3 Nf6 3.e3 c5 4.dxc5 e6 5.b4 {As I said, 5.c4 shoud be playable, but I would have to learn a whole new position type } 5...a5 6.c3 b6 {!} (6...axb4 7.axb4 b6 {gives White a chance to create two connected passed pawns with} 8.Bb5+ {soon to be followed by b5 - this is the only line mentioned in "Killer"}) (6...Nc6 {??} 7.b5 {followed by 8.c6 is moronic play by Black of course, but it illustrates why White cannot play this way in the "real" Noteboom.})7.a4 bxc5 8.b5 Bd6 9.c4 0-0 10.Nc3 Bb7 11.Be2 Nbd7 12.Ba3 Qc7 13.Rc1 Rfe8 14.h3 Rad8 15.cxd5 exd5 16.0-0 Ne5 
17.b6 Nxf3+ 18.Bxf3 Qxb6 19.Rb1 Qc6 20.Qb3 Ba8 21.Qb6 Qxb6 22.Rxb6 Be5 23.Ne2 c4 24.Rb5 Bc6 
25.Rxa5 Ra8 26.Rc5 Bxa4 {?} (26...Rec8 {! followed by ...Nd7 traps the rook on c5 and should win for Black.}) 27.Bxd5 Nxd5 28.Rxd5 Bc6 29.Rc5 Be4 30.Bb4 Bd3 31.Re1 Bd6 1/2

P.S. 
Some years ago I stumbled upon an interesting way to deal with an even earlier ...c5 by Black, again helped by som "reversed" logic. I don't garantee an advantage, but perhaps someone else will want to try it out?

1.d4 d5 2.Nf3 c5 3.dxc5 Nc6 and now the idea is 4.Nc3!? e6 5.e4!? I took this idea from the Chigorin Defence (1.d4 d5 2.c4 Nc6), which i used to play. Black's early ...e5-break is a crucial part of the strategy there. 5... d4 6.Ne2 Bxc5 7.a3 a5 8.Ng3 e5 
9.Bc4 In the Chigorin this bishop often goes to d6/d3, but I thought this looked more active. Nf6 10.0-0 Rb8 11.Qe2 h6 12.Kh1 Qe7 This position, apart from White's extra tempo, could well have arisen from a Chigorin. My game with Lars E..Andersen, Gausdal 2000, continued 13.Nh4 g6 and now the incomprehensible retreat 14.Nf3?, after which Black seized the initiative, though I managed to draw in the end. I think White should open the position with the intended 14.f4, exposing Black's vulnerable king in the centre. 

-Stig K. Martinsen