Of course this is true, as far as it goes.

But I'm inferring something from it in context that you perhaps don't mean, and I'd like to clarify. (Assuming you even read comments from this far back.)

Specific example: a couple of months after you posted this, I suffered a brain aneurysm that significantly impaired my working memory, to the point where even elementary logic problems -- the sort that currently would barely register as problems that needed solving in the first place - required me to laboriously work them out with paper and pen. (Well, marker... my fine motor control was shot, also.)

The question arises: could I have passed this initiation ceremony?

I certainly could not have given the right answer. It would have been a challenge to repeat the problem, let alone solve it, in a verbal examination. My reply would in fact have been "I'm not sure. May I have a pen and paper?"

If the guide replies more or less as you do here, then I fail.

I draw attention to two possibilities in that scenario:

(A) This is a legitimate test of rationality, and I failed it. I simply was less rational while brain-damaged, even though it didn't seem that way to me. That sort of thing does happen, after all.

(B) This test is confounding rationality with the ability to do mental arithmetic reliably. I was no less rational then than I am now.

If A is true, then you and the guide would be correct to exclude me from the club, and all is as it should be.

But if B is true, doing so would be an error. Not because it wasn't fair, or wasn't my fault, or anything like that, but because you'd be trying to build a group of rationalists while in fact excluding rationalists based on an irrelevant criterion.

Now, perhaps the error is negligible. Maybe the Bayesian Conspiracy will collect enough of the most rational minds of each generation that it's not worth giving up the benefits of in-group formation to attract the remainder.

OTOH, maybe not... in which case the Bayesian Conspiracy is on the wrong track.