I don't think that is correct according the to thesis (and again, haven't read it so may be misrepresenting).
They are saying that at that (pro) level, not making mistakes is not enough, you have to make things happen too. Obviously making too many mistakes will cost you, but they are suggesting that playing "perfectly safe" can still easily lead to a loss. I don't think anyone was suggesting that pros don't make mistakes (hell, there are entire stats dedicated to this), but rather than merely focusing on minimizing them will not win you a championship.
The idea is that conversely at lower levels of performance, this isn't the case, and it really is mostly about lowering mistake rates. More tellingly, it suggests that the opportunity cost of focusing on anything else is likely too high.
To tie it back to the actual post, this suggests if you are spending a bunch of time and effort "metagaming" but you haven't already mastered the game, you are probably going to have poorer returns than just getting better at the game. This at least sounds plausible to me.
still, that's just a matter of definition and perspective. even at a pro level, if you don't win, you made at least one mistake relative to your opponent. so to win, make fewer mistakes, right?
i don't play video games anymore (and never played more than recreationally), but i've played pro-am basketball, and from my observation, being a pro doesn't mean you don't make mistakes. it usually means that, besides having highly honed skills, you have more tricks in the bag to recover from your mistakes (and conversely, to induce mistakes in opponents).
so the contention is that everyone can benefit from a strategy of reducing mistakes (aka active practice), pro or not, and the prior anecdote isn't really super-useful.
--devli's advocate