Preferably comparing the time it takes to compile a system with -Os -02 and -O3 compare the disk space they take up and the responsiveness of the system, memory usage etc. I'd be interested to see the differences
Or use -O2 and compare different versions of GCC
Or use the same version of GCC and compare CPU schedulers or kernel versions
Or have a system that's always running the latest Xorg components and latest next-drm kernel doing a compile test cycle effectivly monitoring git for regressions
I would like to cast my vote for Gentoo benchmarks as well. There are two ways to take this of which I can imagine at the moment. The first is a shootout between rolling distributions and static distributions. the second is a comparison of Ubuntu and Gentoo with major software (e.g. graphics drivers, X, kernel, etcetera) set to the same major version, and all other software being Gentoo's default software.
It probably would be a good idea to include two Gentoos in such comparisons, because Gentoo has both a stable tree and a testing tree. The stable tree is what people new to Gentoo usually use while people familar with Gentoo use the testing tree. Both are stable software according to the upstream developers. The distinction is that the package maintainers are being cautious in that they let the more experienced users dog-food the newly released stable software from upstream before the less advanced users are given them.
I would want to see Ubuntu 6.06 with the FGLRX driver on an ATI ~R280 (radeon 9200 or so) against Ubuntu 10.04 with the open source driver. I untested one of these card recently and found the 3D performances as good as with the old proprietary driver. I want to see if the open driver made that much progress or the games just got optimized.