Perhaps. I'm a little concerned that if this approach is used for diagnostics that it becomes another form of gatekeeping. (i.e. someone is transgender but denied treatment because their brain scans come up as cis.)

There's a dark side of everything, but there are people who believe they want to transition who then find out it doesn't help when they do. Being able to diagnostically assess the likelihood that it's the correct treatment would be good for everyone.

Conversely it could help someone who is in denial and experiencing disphoria to come to terms with themselves.

That would have to be an incredibly precise test though; if someone is in denial and you tell them "This test says there's only a 5% chance you're right", they will probably stay in denial.

Reminds me of XKCD's Significant comic: https://xkcd.com/882/

So, that gets complicated. What if medicine finds a certain physical trait that is prevalent in 90% of cases where people are trans? What does that mean for the other 10%? Odds are good their lives just got WAY harder. Coupled with a history of being mistreated by the medical industry, there's quite a lot of (rational, imo) fear in the trans community about medical diagnoses.