I have recently won a funny game (as white) in a well-known theoretical line of the Qc2 Nimzoindian. White's pawn sacrifice is quite dangerous, and after a couple of inaccuracies Black ended up resigning in a position where he is in a zugzwang situation. However, I think white has nothing if Black opts for the less well known, but safer 14...Nd7 (instead of 14...Bd6).

The best source to learn playing the Nimzo is the old landmark book by Gligoric, no matter how outdated the included theory is. If you can find a copy anywhere, just grab it. It is probably the best opening book ever written: The great Glika was always extremely objective in his evaluations (always playing against pieces, not opponents, and this just shows everywhere in the book.

I just won a game as Black in the aforementioned 4.f3 variation.

It was a "lucky" win (if you can call a win in a correspondence game as lucky!), factly I was stunned when my opponent sacrificed his knight, when 23.Nb4 would be flat equal (factly everything up to there has been played before in another correspondence game). In the post mortem, my opponent explained that when analysing, he had mistakenly put the other rook on b8!

Study the typical pawn structures... I suggest to you the idea of playing Bb4 and play with d6, Qe7 and later e5 the idea is to exchange the Dark Square Bishop and put the pawn on the dark squares because all your remaining minor pieces should be very good. I personally doesnt like the variation used by pfren not because its bad a lot of top players use it but its a matter of tastes.

Interesting? Hmmm, I find it slightly dull.

After the accurate 15...Na6! Black has zero problems, but he could easily fall in the positional trap 15...Be6? 16.Nd4! and white has fantastic strategical compensation if Black takes the piece (although he has nothing better than that).

Typical theme of that variation! 16.Nd4!! Moskalenko write a book about opening and talk a lot about the Nimzoindian more specifically that variation.