I agree with linux4all and Kano in these two topics.
Honestly, I prefer benchmarks on hardware so using the same distro measured across slightly different hardware configurations or just a benchmark on a particular hardware component.
It seems distros vary too much for benchmarks and there's so many variations in the packages and the versions. It seems to be a very subjective viewpoint and although there's benchmarks to examine here, there doesn't seem to be too many deal breakers unless you're partial towards a certain distribution. It's a matter of personal taste, to put it simply?
I'm not an Ubuntu fanatic nor do I preach to use Ubuntu over others but I use it because I'm not adept enough at something more complicated or more hands-on like Slackware and Gentoo (related distros). Also, I started with Debian and prefer to improve my knowledge and expertise (lol) with it or debian-based distros rather than jump into something significantly different.
Yeah, Linux is Linux but imho, better to excel in one than be somewhat proficient in a few or several. However, I do like Red Hat and Fedora and I think it's good to have considerable exposure in at least one of those.
Given a choice, I prefer benchmarks between video cards/drivers, for e.g., over distro comparisons.
I doubt I'd be able to tell the difference between distros much, anyway, so if there are any benchmarks to provide vital info, I think comparisons of desktop environments would prove more useful. Perhaps, a measurement or comparison of how well distros integrate DEs would be interesting, too?