For web browsing and playing video (i.e. 99% of all users' "daily driving"), I don't think it even matters to consider these sorts of benchmarks.
I do edit video and work with 3D graphics almost daily though, so this is sort of comparison is very relevant.
In other words, of you were to actually measure page load speed, are you saying the measirements would correlate with the benchmarks?
In reality I think for general web browsing the difference, if it exists at all, in either direction, would not be perceptible.
I know that's not EXACTLY what you're asking, but I think it's worth mentioning the practical aspect as well.