It's been a while since I saw it, but they weren't even doing Chess960. They were actually randomly placing pieces on the board, often in configurations that never occur in a real game.

The other part I'm wondering about now is power. Even if there were no statistically significant advantage found in the study, that doesn't mean there's not enough real advantage to see a difference over the course of a tournament.

Undisciplined

Might be a different effect, 1-50, 50-99th percentilen versus the top 1 or top 0.001% 

denlillaapan

Stacker_Sports

Magnus Carlsen On Why the Future of Chess Lies in Freestyle (The Economist)

I wonder what level of players that study used though, because a fascinating thing about Chess960 tournaments is that a lot of the best classical players still end up at the top.  