[tomsto]

This.

I wonder if a big part of the problem is that depression and other mental illnesses are slightly unusual in that they are categorised based on symptoms rather than underlying biology, which are poorly understood. As a diagnosis it’s about as useful as ‘chest pains’

Maybe (I am not a clinical psychiatrist) there are 1, 10, or 100 very different underlying causes which would indicate different treatment if only they were better understood. But nonetheless the existing classification / ontology is established and will be hard to displace.