@r.b: 

you ought to subscribe to chesspub. please see recent chesspub updates or any recent issue of New in Chess. 5. e4 is not a harmless move (as was usually thought to be), but a newly fashionable and quite dangerous weapon recently used by ivanchuk, karpov, aronian and sokolov. I won't pretend it's the best anti-nimzo weapon. only wanted to make the point that you can't guarantee a Ne4/f5 setup in the nimzo, as you seem so insistent upon.

ah yes. how can i forget 8... f5 is a critical line in the 4. f3 system  But f5 in that line is hardly meant as a king side attack, which is presumably what is wanted here.. and yes there is no Ne4.

Incidentally the classical dutch setup with 1. d4 e6 2. c4 f5 3. Nc3 Nf6 is a good way to go, though a successful kingside attack needs a co operative White. Nowadays i sometimes play the "rubinstein-style" 4. e3 (recommended by Palliser), instead of 4. f3!?, even though the lines frequently merge.

6..Ne4 is playable although on the basis of latest theory white seems to have a tiny edge.  But what about 4. Qc2 0-0 5. e4! No f5 here iirc. Even if you do get it in, you're hardly well-placed to attack k-side. 

not to mention the 4. f3 nimzo where in 1000+ games (blitz or otherwise) i have not seen anybody play f5. i repeat: you cannot always get what you want as black.

That is why I play f5 on the 2nd move.
Still the conclusion of Inn2 is slightly wrong. Not because Black misses his king's bishop it is hard to play e6-e5, but because Black has exchanged it against the White knight.
If both black squared bishops are from the board, it is much easier to force e6-e5. I also think, that leaving the bishop on c8 (protects e6 and f5) helps even better.

When Black combines a queen's fianchetto with f5/Ne4, he should consider c7-c5. Another option is pieceplay as indicated by Inn2 - Rf8-f6-h6. But I agree, that this is only succesful if White cooperates. Williams in his book on the Classical Dutch gives two examples, but not in the theoretical part.

Spielmann has played the Dutch with Bb4 a few times, with reasonable succes.
Rubinstein-Spielmann, Oostende 1907: Black was lucky to escape with a draw.
Rubinstein-Spielmann, Prague 1908: White won.
In both cases Black fianchettoed the queen's bishop.
Rubinstein-Spielmann, San Sebastian 1912: Black won a great attacking game - but the queen's bishop went to d7.

@Inn2

where is the problem in getting in Ne4/f5 in the 4.Qc2-NI ?

4. Qc2  Bxc3
5.Qxc3 Se4
6.Qc2  f5

havent looked at 4.a3 in the QI, but I think also there you can get in Ne4/f5.

i don't understand this obssession with f5. you guys have crude taste  so do i sometimes, but as black you cannot always get what you want. 

i do not think it is possible to succeed with a k-side attack in most variations of the ni/qi... assuming you can even get in f5 in the first place (can't do that against the most popular variations e.g 4. Qc2 NI or petrosian QI). you'll only succed in your "auto-pilot" Ne4/f5/Rf6/Rg6(h6) if your opponent's play is terribly cooperative!

"if after Bd3 you overprotect Ne4 with d5"
In general this is a foolish thing to do. One sample line:
1.d4 f5 (sorry, I am too lazy to find out how it works in the Queen's Indian) 2.c4 e6 3.Nc3 Nf6 4.e3 b6 5.Ld3 Bb7 6.Nf3 Bb4 7.Bd2 o-o 8.Qc2 a5 9.a3 Bxc3 10.Bxc3 Ne4 Flohr-Bondarevsky, CCCPch 1951.

by the way: if after Bd3 you overprotect Ne4 with d5, then - I suppose - you have transposed to the Dutch Stonewall, havent you?

Well, an interesting idea, to play a Dutch, but start it as a Queensindian. 

This would keep all those maverick attackers away who bother me in my Dutch with moves like 
h4    
Nh3 
c3+Qb3 
0-0-0
2. e4 
2. Bg5 
2. Nc3
2. Nf3

Why not turn it the other way round?
1.d4 e6 (or f5 first)
A) 2.c4 f5 3.g3 (3.Nc3 Nf6 and both Bb4 and b6 are common moves) Nf6 4.Bg2 Bb4+ 5.Nd2 (5.Bd2) o-o 6.Nf3 b6!? idea 7.Ne5 Ne4.
B) 2.Nf3 f5 3.g3 b6 4.Bg2 Bb7 5.o-o (5.c4 Bb4+) Nf6 6.c4
with the sort of setups you are looking for. In line B 7.d5 though is a serious threat.