EAX is more than just reverb BTW. It also models occlusion, obstruction, air absorption and a whole bunch of other things.
Why is it that people do not like EAX? If used well, it can greatly enhance the game experience. I suppose though, that if it is used poorly it can be a detriment. However, if it IS used poorly, that is the fault of the game developer.
You can think of 3D audio as having 2 parts. The first is taking a sound, figuring out how it will propagate based on game geometry, and expressing this propagation as a set of filters applied to the original sound. The second part is to actually apply the filters to the sounds. EAX does the second part of these 2 steps. A3D did both. The problem with A3D was that what it did for the 1st step was an approximation of a VERY complex problem, and it was locked down in hardware that would not have necessarily scaled to support today's game environments. Also, the implementation of the 2nd part was necessarily going to suffer because you can only have so much hardware. Creative decided that since the 2nd part is relatively straighforward to accelerate in hardware, they sould concentrate on that, and leave the 1st part to the CPU. After all, there are a lot of tricks and optimizations that you can do on the 1st part if you know specific things about the game. But these tricks and algorithms are still evolving and will vary in applicability from game to game and engine to engine. A3D was kind of brute force.
You can argue that EAX filters degrades audio quality, but any kind of processing is going to do that. The new X-Fi will hopefully be cleaner because of its floating-point DSP. You can argue that perspetive corrected texture mapping of 3D polygons messes with the image quaility of the original texture images (that's why we do billinear, triliear, anisotrophic filtering, etc.), but that is a tradeoff we are willing to make to get cool 3D graphics.