> performance given a certain baseline training volume

That would make the argument pretty circular, because of course those two tiny slivers of the population have high overlap. I was trying to assume good faith, despite the venue, and clearly that was a mistake.