I prefer the word "claim" - it more useful and accurate. ... There is literally nothing to distinguish a TW from a man claiming a TW identity, because they are by definition the same thing.
Stongly agree.
Which in turn means the population of "transwomen" will be different depending on environment, and who's making the claim.
In Iran, men claim to be women because homosexuality is outlawed, so "transwomen" will tend to be homosexual.
In South America, men perform as women in prostitution, so "transwomen" tend to be prostitutes and at high risk of violent crime. (This is where the bulk of the "transwomen" being vulnerable comes from.)
In the UK, an awful lot of sex offending men claim to be women after conviction, so in prison "transwomen" tend to be heterosexual sex offenders.
Any statement or data about "transwomen" is going to be very dependent where you found your "transwomen" - it's not going to be generally applicable to "transwomen" in some other scenario, as that will be a different population of men who trans identified for different reasons.
So going back to your original post - there is data that "transwomen" in prison are disproportionately sex offenders, but there's no way you can assume that's representative of non-prison "transwomen".