> I've yet to see any evidence that it actually does reduce bad hires
And the OP's own point "responsibility is diluted so if there's a bad hire you can blame the process rather than one manager" would be needless if the process actually did avoid bad hires.
It's not a reasoning I would agree with either. The core of my original comment is "avoid responsibility": people don't want to be held responsible for a bad hire (or a discrimination lawsuit), so adding multiple rounds to the process shares that responsibility among as many other people as possible. By cargo-culting from FAANG you can also use the "nobody got fired for buying IBM" argument and this dilutes responsibility further.
So does that actually reduce bad hires (or increase good to excellent hires)? Probably not, but that's not really the purpose of all this theatre. The purpose is that if there's a bad hire, do I, personally, get the blame?
And the OP's own point "responsibility is diluted so if there's a bad hire you can blame the process rather than one manager" would be needless if the process actually did avoid bad hires.