Not even 24 hours after the G5 was announced, people around the web have started refuting Apple's claims on the apparent speed of their new flagship computer. Oh well, these people just haven't learnt not to trust ANY benchmarks put out by computer-makers. I dare you to find me a page on AMD or Intel's web site that shows the benchmark performance of their processors performing poorer than the competition. It's all in the marketing, folks!
You would have thought people have learned that from the nVidia and 3DMark2003 'driver optimization' fiasco (yes ATI was guilty of that sometime in the past too). I don't find artificial benchmarks a reliable indication of performance. Even if people really want to see benchmark scores, they would go to independent hardware sites like Tom's Hardware or AnandTech for an objective and fair comparison, not Apple or Intel or AMD. The thing is, the PowerMac G5 is not shipping yet, and such independent tests are not possible (not even sure HOW it's possible, being totally different CPU and system architectures), so what's the point of nit-picking on the figures posted by Apple?