I'm going to try to look at this from a fanATIc's POV just to keep this fair, and possibly someone more knowledgable than myself can answer in case a fanATIc does bring this up
From anandtech's review SM3 only seems to benefit special circumstances. It would be interesting to see a line graph showing the FPS before and after SM3.
Has the average FPS increased simply because only a few seconds of the benchmark peaked much higher than without sm3, but the rest of the benchmark stays relatively the same?
ChrisRay - Does the rest of the game "seem" faster?
Hopefully tomorrow some more benchmarks come out that answer these questions. I really would like to see nvidia shove SM3 down ATI's throat to be honest

ATI has had bragging rights way too long.