Thanks for the replies, I am going through them today. I noticed I had cut (or rather pasted) short a variation.
In stead of 1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 Be7 I meant 1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 Nf6 (or some other move order) 5. Bg5 Be7 (or with h6 Bh4 included).

1.d4 d5 2.c4 e6 3.Nc3 dxc4 4.e4 is pretty much just better for White, though I guess Black can play this way if he wants to.  At best, Black will reach a line that usually comes about after 1.d4 d5 2.c4 dc 3.e4 c5 4.d5.  Nowadays, 3...c5 is considered inferior to 3...Nc6 and 3...e5, though it's certainly playable. I refer you to theoretical sources for info on this line.  If Black avoide 4...c5 he will just be clearly worse.

1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 Be7 is a line that I've never come across in play.  I might personally prefer to try to steer the play to a Closed-Catalan type position, though a simple way of playing might just be 5.e3!?.  Now when ...Nbd7 comes Black will just have a worse form of a Meran Semi-Slav, as the Bishop is on e7 instead of d6.  You might even try Shabalov-like play, i.e.: 1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 Be7 5.e3 0-0 6.Qc2 Nbd7 7.g4 (or something like that) is probably a good pawn sac.  If 5.e3 isn't to your taste, you might try 5.Bf4 when Black is denied the "standard" line with an early c7-c5 push.  However, Black might be able to take advantage of this particular move order with something like 5.Bf4 dc.  Personally, I'd play either 5.e3 or try to get a Closed-Catalan with 5.g3 or the like.  By the way, there are many good ways for White to play the Meran without resorting to the extremely trendy Shabalov stuff.

Against the Slav, 3.Nc3 or 3.Nf3 is purely a matter of taste.  Traditionally, 3.Nf3 has been preferred to avoid the Winawer Counter-Gambit (3.Nc3 e5) and 3.Nc3 dxc4.  However, I personally like playing against 3...dxc4 and the Winawer, so I play 3.Nc3.  It's a matter of taste, but if your intention is to play the main-line Slav positions anyway, it's considered better to play 3.Nf3 because it avoids Black's 3rd move deviations.  After 1.d4 d5 2.c4 c6 3.Nc3 Nf6, I play 4.e3, which is a whole different story...

That Korchnoi-Kasparov game is a strange one; as far as I know, the only way for White to get anything after 1.d4 d5 2.c4 c6 3.Nc3 dc is with 4.e4, when something like 4...b5 5.a4 b4 6.Ne2 Nf6 7.Ng3 can prove quite fun for the first player.  After 4.e3, I always thought the recommended antitode was just 4...e5 and Black is already equal.  But then, those players know a lot more about chess than me!


edit: Wow, while I was typing the above post two people already responded.  I hope that helps!

I'm no QG expert and I'm also working with the Cox book at the moment. I still try to answer.

1) 1.d4 d5 2.c4 e6 3.Nc3 dxc4 is a quite rare line. I've looked a few references which dismiss it quickly. The second revised QGA book by Sakaev and Semkov treated it more seriously. It's given a full chapter. (5 pages)

1.d4 d5 2.c4 dxc4 3.e4 c5?! 4.d5 e6 5.Nc3 exd5 6.Nxd5 Ne7 7.Bxc4 Nxd5 8.Bxd5 Be7. 9.Ne2!? seems to be their preference although 9.Nf3 also looks okay.

2) 1.d4 d5 2.c4 e6 3.Nc3 c6 4.Nf3 Be7? No. Cox prefers 4.e4. After 1.d4 d5 2.c4 c6 3.Nf3 Nf6 4. Nc3 e6 5.Bg5 Be7 is considered passive because c6 is played too early. See http://www.chesspub.com/cgi-bin/yabb2/YaBB.pl?num=1147919525

3) 3. Nc3 is as good as 3.Nf3 and is played quite often nowadays. One reason for playing Nf3 could be that the e5 gambit is ruled out. (A reason for Nc3 is not allowing a proper slav at all with 3...Nf6 4.e3 or 3...dxc4 4.e4.)