Most of all CellFish (or DellFish) was confusing a lot of different aspects of computing:
1) The OS
2) The programs
3) The hardware
4) The marketing
You cannot effectively compare the efficiency of OS'es if they are incapable of running on the same hardware. Thus all you have to rely on is subjective experience. The same more or less applies to programs: even when running the same programs (or benchmarks) on different harware/OS's you cannot tell objectively which setup is better.
Hardwarewise, matters are equally complicated, since a dual G4 is very different from a pentium 4 in many respects, programs will perform differently and behave differently: again what is better? Well more speed is better, but personal experience is no scientific argument. CellFish says OS X is slow, I say OS X is fast. We could be both right, since our setup, programs, etc differ.
Then you come to marketing and you enter a very dim lighted area: which arguments used by commercial companies are really good? Which benchmarks / revieuws / test can you trust? Everyone will say his product is best.
So really there are at least four different discussions going on:
-Windows against the Mac OS
-Apps under Mac/G4 vs. Win/pentium
-The G4 (+mobo, FSB, dual etc.) VS the Pentium as hardware architectures.
-The reasons companies use to market their products and how they perform in the market.