Who's fudging what, then? I think they're miffed over the Intel changeover.
If you read the comments on the original article, the commentators point out MacWorld's statistical approach is backwards... for example: if the speed is increased from 100 to 150, then it is either a 50% increase or a 33% increase, depending on whether you are looking at the starting speed or the increased speed.
For example Jobs stated that one application was 1.84 TIMES as fast on the new machine compared to the old... that is either a 54% increase... or 84% faster... almost twice as fast... or only 1/3 faster. Depends on perspective.