if we stop thinking about the fact that as an abstract, general question a random human being is much more likely to be cis than trans
That said, it could also be taken as advising you not to double-count your priors by using them to discount the evidence. Imagine you've drawn a ball from an urn, and the ball looks blue to you — but your priors say that 99% of the balls in that urn are red. How much time do you want to spend questioning the validity of your color vision or the lighting before you consider that you drew a rare ball?
Is cis or trans identity really something that is truth-apt (& therefore in the purview of probability)? It seems to be a combination of self-description of feelings, plus chosen group affiliation.
The self-description of feelings is presumably more or less infallible, and the group affiliation is stipulated by the individual.
Rationality quotes time!
The usual rules: