Considering we have many dozens of distros and ten of thousands of their permutations (different "spins", desktop environments, etc.) this test is always going to be distro specific.
I chose Fedora because of its freshness, proximity to the source (Fedora prefers to apply as few patches as possible vs. e.g. Debian/Ubuntu) and readiness. You just install it and start working.
Fedora also bundles a bunch of software by default for a nice desktop experience for new users that you wouldn’t get with only Gnome on Arch. The difference is quite big, I’ve used both and Arch is way leaner even with a full install of the Gnome desktop.
True so yet all of them have the same/similar background services/daemons/applications, so the comparison is still valid though it will be distro-specific.
I cannot physically test all the distros and their permutations, and then there are some user-defined ones such as Gentoo or LFS.
Now do it again on a different distro which packages different default apps and services, and see how these values differ by multiple hundreds of MB.