the controllers were on numerous older boards with raid (dont remember specifics), a couple nforce2s, an nforce4, promise fasttrak s150 tx4 (only a 33mhz pci card, so 2 raptors already saturated it), recent ecs intel chipset (dont know the controller), different stripe and cluster sizes... but they all were pretty similar, both on and off of raid... at least where it mattered (windows boot times aside, how often do you really need to restart though), i will say windows is slightly more responsive though in raid 0, so thats a plus
as far as the older raptors sucking... no command cueing, 8mb cache, 36GB platter, GD revision, hotter noisier more power hungry running in raid... newer one ncq, 16mb cache, 74GB platter, ADFD revision, quieter, and cooler, less power hungry (because it wasnt running in raid)... they were 3-4 years older, but even with 4 of them being in raid, it should make some difference above the single faster raptor... too many other articles pointing to the contrary though for average uses (the average user)
i used to be a heavy advocate of raid 0 (for everything disk related even "you want faster, just go raid, because the benchmarks say so")... up until i started paying attention to it, dissecting it beyond just benchmarks... to find out how much i was imagining things, and wasnt... raid can help (no doubt there), but for certain things it just really doesnt... and for those things, you are simply better off going without it... the average user for example, wont be able to see much difference at all, even for gaming (its very game specific at that point too though)... games arent designed for raid, some games see no benefit, or very little, current FPS show very little benefit for example (i know enough people want to claim the contrary to that though)... strategy games however may show a large benefit to loading (completely different game design)