Suppose that these anecdotes do generalize and are in fact representative. (That's a big supposition!) 1/

Then if we were to develop direct measures of capability rather than relying on sex as a correlate, we'd end up effectively reconstructing sex-segregated sports, EXCEPT we'd have an ability to place outliers, including spontaneous outliers like Caster Semenya and new outliers like trans people. 2/

Maybe it's a long-term project, and in the meantime we have some muddling through to continue. But why would it not be a desirable project? /fin