With regard to lumping disorders together, this is the camp that I fall in. I think we will look back at our present diagnoses the same way we look at a historical diagnosis of "fever;" it's correct but not as precise as it needs to be. The prevailing direction in research is to move toward clusters of symptoms rather than diagnostic labels. The relatively new NIMH RDoC will help with this:
https://www.nimh.nih.gov/news/science-news/2015/the-nimh-res...
One of the other things I'm seeing a lot in this thread has to do with the tension between autonomy and paternalism. This is a problem that keeps me and many of my colleagues up at night. At what point do you step in to limit a person's decisions? That's not something that most of us take lightly. I sure don't. At this point, in the US, the primary criterion for making decisions for someone is based on preventing injury to themselves or other people. This is a reasonable, but complex, dividing line. An unintended consequence is that it continues to over-emphasize and reinforce the connection between mental illness and violence.