I found them quite different, actually. GMATPrep's questions were much harder (especially in quant) but their scoring was much more leniant. In contrast, PP's questions seemed much more like the "real thing" (this year, at least) and were much easier, but the scoring seemed tougher. I think I got a lot more questions wrong on GMATPrep's CATs but still scored around the same as I did on PowerPrep. Both scores turned out to be pretty predictive of the real test, though the level of difficulty of the questions was much more like PowerPrep in my opinion.