Obviously, even bottle necked by a really slow bus, a more powerful GPU should perform better.
It just seems wasteful to place further limits on a decent GPU's performance.
Acube's use of slow, relatively low power PPCs without AltiVec already limits performance.
While the X1000 supports AltiVec (and is quite faster), the X5000 won't and at 2.2 GHz it has a real speed disadvantage when compared to the X86 systems that modern video cards are usually run in.
While all these developments ARE improvements to our environment, we still operate at a disadvantage when compared to mainstream systems.
This, when combined with uncompetitive pricing, discourages further expansion of our community.