Agreed on both counts. Hormone receptors are genetic: sex is what the whole genetic programme does, acting over months, through gonads, hormones, receptors and tissues. “Genes determine sex” and “sex develops” aren’t rivals; the second is how the first happens. And yes, observation at birth is highly accurate — as a prediction of how that development will go, because the dimensions usually line up. An accurate proxy is still a proxy. The cases we’re arguing about are exactly the ones where the dimensions don’t line up, and there the proxy is the thing that fails.
As for gametes: around one man in a hundred never produces a sperm in his life, and some women never produce a viable egg — and nobody hesitates to call them male and female. So “defined by gametes” has to mean “organised to produce, under typical development,” which is a judgement about what the body is for, not an observation of what it does.
It’s also a definition that philosophers introduced for classifying people about five years ago, after chromosomes and genitals kept failing — so if you want to talk about who’s been changing definitions...
And btw people don't check a baby’s gametes in assigning sex; the birth certificate records anatomy, and anatomy is being used to infer the gamete type. The proxy is doing the work, and the proxy is what fails in the cases we’re arguing about.