Ah, but it's stronger evidence that your expectation is wrong

Keyword stronger. The claim you were questioning was whether there was evidence at all. I do nothing more than support the claim that it is evidence.

and self-reflective priors would have 'expectation is wrong' starting more likely than 'interference from an outside agency'.

Probably, given roughly human-like intelligence with information roughly like what we have now. The counterfactual wasn't specific in that regard but did suggest an assumption of a particularly strong understanding of human nature.