@GriseldaandMike
"Men aren't stronger than women just because they have testosterone in their system, they are stronger than women because testosterone makes them develop differently."
You are absolutely correct. And it brings us back to the friction point from the essay: the difference between a categorical biological boundary and an actuarial risk assessment.
I perceive your argument as stating that if a medically transitioned person does not achieve full biological equivalence with a natal female down to the bone structure, they must therefore represent a baseline physical risk significantly similar to a testosterone-dominant male.
But in actuarial math, risk is more like a gradient.
To use your "lever" analogy: a lever is only as effective as the kinetic force pushing it. While the limb length remains permanent, the muscle mass required to operate that lever at peak male force diminishes significantly without testosterone, as does the male-pattern aggression that often drives violence.
The physical risk profile of someone with male skeletal levers but suppressed testosterone is fundamentally different from both a natal female and a baseline male.
It is its own distinct actuarial category.
If your standard for a space is that an individual must have zero male-developed physiological traits, that is a perfectly rational categorical boundary. (And it is exactly why sports categories are debated, although personally, I think the current arguments are useless, since the reason why trans women were banned from playing women at chess still apply to other sports, yet, hopefully, we all can agree men are not biologically better than women at chess).
The essay's core observation is that we have to be honest about our metrics.
By requiring full biological equivalence rather than measuring the altered, reduced risk profile of the suppressed individual, that is enforcing a categorical rule, not calculating an actuarial equation.