Mostly I want to say: There's a ton of evidence that covid damages cognition to some degree, even in clinically mild cases. But the specific relationship between subjective experience and objective measures is genuinely complex, so it's good not to over-reduce it.
(Metacognition is wickedly tricky, as anyone with a history of "brain fog" and/or loved ones with cognitive decline can attest. Some conditions are subjectively detectable and some not and severity CAN play a role but doesn't always.)