Maybe they could choose anatomies. But they can't choose the part they play in reproduction later. If they choose an anatomy discordant with their genetic gender, it can make reproduction tricky.
More seriously, I'd say their personal preference should be the most important. It's not like we have a shortage of people.
Or is the issue that with male chromosomes it's "tricky" to get working ovaries?
But what if it's not the typical transgender surgery, but actually having the gonads and tubercle turn into ovaries and clitoris instead of testicles and penis?
I imagine it would produce about half X-ovums and half Y-ovums. X-ovums should be normal, Y-ovums might not be viable at all.
If they were, Y-spermatozoons could probably not fertilise them, or in any case it would fail rather fast without developing into a viable embryo. X-spermatozoons could fertilise them to produce male embryos.
Producing working ovaries with male chromosomes is the most unlikely step, but otherwise it seems like it should work provided the biological machinery is there to support pregnancy.
That said, the fact that one gene seems to have so much effect suggests it's plausible that in the future we might be able to do something to make those individuals fertile, or to enable genital reconstruction surgeries for transgender people that enable new fertility.
Note that much of this may be oversimplified, so any experts can feel free to apply some pedantry to the finer details.
[1] https://en.wikipedia.org/wiki/Testis-determining_factor [2] https://en.wikipedia.org/wiki/XX_male_syndrome [3] https://en.wikipedia.org/wiki/XY_gonadal_dysgenesis
https://www.google.com/amp/s/mobile.nytimes.com/2017/05/16/h...