It's sad that your bar for evidence is so low. I would like to see more research in this area and am in no way trying to argue against it. More research would in my opinion be useful but it's not this.
In medical reseach there are several red flags for poor research and this one has too many for me to be comfortable with it being considered as evidence.
There were only 24 brain scans. That's not enough for the normal analysis needed
It mentions other studies where the same effect wasn't found. Then tries to explain these away with no evidence that this is the case
They used a "recently developed multivarient classifier" so their methods are new and novel and untested. They are not standard methods with a long track record in a larger body of research
It's not a replicated study. It's a small study, using novel methods and not been tested at other clinics on a larger population
There is a lot of loaded language thoughout i.e biological sex, perceived gender. Perceived? Very few biological men look like biological women even after drugs, surgery etc.
I think trans people deserve better. Research like this doesn't help.