First we reject this argument as a naturally-occurring brain that is sufficiently-trustworthy-for-survival (per Natural Selection) is not equivalent to a brain malfunctioning under the influence of hallucinations.
I would also point out that the existence of (occasional) hallucinations has less in common with the perfect trustworthiness that one would expect from a perfect Creator than with the only approximately trustworthy result we would expect from Natural Selection.
That a reasonably trustworthy perception of reality has fitness benefits should be obvious to anybody who is not an apologist.
And if the moon were made of green cheese, we’d be able to mine it for dairy products. Your hypothetical is no more substantive than this one.
And I have only been specifying that Natural Selection would produce a reasonably trustworthy mind – and had in fact explicitly stated that cognitive biases exist. It does not seem at all likely that a propensity to believe in UFOs would have a sufficient fitness penalty that Natural Selection would work to eliminate it. And even if it did, the idea hasn’t been around for long enough for Natural Selection to go to work on it.
I saw a similar, and similarly-ludicrous, argument from Alvin Plantinga in defense of his execrable Evolutionary Argument Against Naturalism (that an Evolutionary argument, from a philosopher whose main source of information on evolution appears to be evolution-denying ID-Creationists, turns out to be bad would seem to be entirely predictable) – from memory it involved running away from tigers in order to make friends with them.
Yes, the protagonists of the Battle of Waterloo are not discoverable by science strictly construed, but are discoverable by analogous evidence-based methodologies – ‘science, broadly-construed’, as some have termed it.
Now such methodologies cannot answer questions like whether Yahweh, Allah, or Vishnu exist – but then such questions do not appear to be even remotely amenable to objective adjudication by any alternative method either.
And even if Science (even ‘broadly construed’) is not universally applicable, does not contradict the point that it acts as a corrective to cognitive biases in the areas where it is applicable. So my original point remains unrebutted.