This “Gettier” attack seems to me to have no more interesting content than a “stopped clock”. To use an extremely similar, extremely common phrase, the New York Times would have been “right for the wrong reasons” to call Scott Alexander a racist. And this would be conceptually identical to pointing out that, I dunno, crazed conspiracy theorists suggested before he was caught that Jeffrey Epstein was part of an extensive paedophile network.
But we see this happen all the time, in fact it’s such a key building block of our daily experience that we have at least two cliches devoted to capturing it.
Perhaps it would be interesting if we were to pick out authentic Gettier cases which are also accusations of some kind, but it seems likely that in any case (i.e. all cases) where an accusation is levelled with complex evidence, the character of justification fails to be the very kind which would generate a Gettier case. Gettier cases cease to function like Gettier cases when there is a swathe of evidence to be assessed, because already our sense of justification is partial and difficult to target with the precision characteristic of unexpected failure - such cases turn out to be just “stopped clocks”. The sense of counter-intuitivity here seems mostly to be generated by the convoluted grammar of your summarising assessment, but this is just an example of bare recursivity, since you’re applying the language of the post to the post itself.
I suppose I must be confused, your saying that the piece was interesting was just because it made you think about the phrase “Gettier attack”?