Unfortunately for their summary, their results show no such thing: they don't even try to robustly group brains as "male" or "female".

The work itself is good, but the reasoning behind their summary is extraordinarily poor. If you want to show that there aren't distinct classes, you at least need to try to classify them. They didn't! They showed that there was substantial diversity (completely unsurprising, though quantifying it is useful), that the diversity overlapped (again completely unsurprising), and then...quit.

Studies that actually try to classify male vs. female brains generally show moderate to considerable success when they do so. Here's a fairly recent paper where they get 93% accuracy--and presumably they could do better with a more exhaustive dataset; this is just what they could do with the data they had, in the imaging modality they chose to focus on, which is far less than the full set of actual relevant biology: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6418873/

So, anyway, the first paper shows that brains are like bodies, except even a bit less distinct: if you look at any one feature, it's not going to be a very reliable way to tell. It does not follow that overall you can't form two classes with only a small number of ambiguous cases--to know that, you have to actually try to do it.