So far those are some stunning numbers, particularly the percentage of folks that actually "hate" 4E (over 22%), and the huge difference in the approval ratings (almost 60% for 3E vs. 46% for 4E).
I think the polls are good, but there is an issue in the "love" and "hate" language.
I voted that I "hate" 4E. That doesn't mean that I want to punch it in the face, any more than the fact that I "love" Classic Traveller means that I want to marry it and make babies inside of it. The language sounds very strident but when I used it I just mean that, having played 4E, I would rather not play any game at all than play that game. Whereas, to me, "dislike" means that I might play it grudgingly or for some other benefit.
I agree that 4E has been very polarizing. I'm not a fan of 3E either (though I voted "Mixed Bag" because the question included the OGL, which I think is a great thing and allows for the creation of very light, Old School games apart from number-crunching dicefests like 3E), but I think that one thing 3E did was open up the D&D hardcore base (us dudes) to more sophisticated thinking about the structure of D&D and how the rules cash out in actual play (and thus produce differing play experiences). Those who have interacted with me online know that I'm an Old School Renaissance guy; but I don't think the OSR could have happened without both the questions that 3E raised (even if it answered them wrong, from my perspective) and the transparency of its design. On top of that, the OGL lets you end up with things like True20 which are pretty light and elegant.
So 3E was more things to more people. But 4E is very focused, very untraditional and very integrated. It is what it is, it does what it does and if you don't like it take a hike. This, coupled with the fact that it does not approach things in the way of its predecessors, and also with the fact that some of the designers evinced a smug and condescending attitude during the runup to the game's release has resulted in a fractured fan base. There was a reactionary phase in which the camps were beginning to form but really neither the supporters nor the detractors really had much of a chance to give it a spin; by now, we've all had a chance to give it a fair shake (hopefully a fair one!) and the fracturing hasn't subsided.
Part of the problem with a rigorously coherent design is that it's less flexible than a sparse or distributed design. In the case of the former, you pretty much are either in the market for what it is selling or you're not; in the latter, there's more room for negotiation.