Interesting. Thx for the post, OP!
I wouldn't call it "scientific", but they go into a lot of detail about the testing and have discussions with the participants about how it feels to switch back and forth between the three setups.
I'd say it's definitely not scientific, given that it's sponsored by a graphics card manufacturer, and they complicated things by putting a different graphics card in each machine. It's unclear whether the issues w/ the 60fps machine were due to framerate alone, or due to frame timing issues as a result of the slower graphics card. And while the testers didn't know which machine they were playing on, the experimenters knew, and Linus is pretty blatantly pro high framerates -- he's cool and all, but I'm not sure that I trust him to be able to set aside his biases, here. On top of that, they were testing in engine, but not testing a networked multiplayer game -- I don't know how much the differences they saw would matter once you add network shenanigans to the mix.
That said, that is a pretty dramatic difference, and I might be wrong in stuff I've said earlier about high refresh displays being a placebo. :-)
Too bad that this isn't something with a lot of academic appeal. It would be neat to have some better design, unsponsored studies to look at. A lot of the tech world runs on gut feels and hearsay; I wish that more of it was based in more rigorous research.