I think there's a lot more D than S in the DSM. I think the DSM is useful to describe the current state of psychiatric thinking and does provide some sort of platform for standardization.

I have a vague recollection that either in the transition from DSM 2 to 3 or 3 to 3R they removed diseases that had an organic rather than psychiatric cause. The implication of that seems to have been that what remained did not have biological underpinnings.

I think that Dr. Insel may be correct in suggesting that reliance on the DSM may be unduly hampering mental health research. The domains that he suggests: Negative Valence Systems; Positive Valence Systems; Cognitive Systems; Systems for Social Processes; Arousal/Modulatory Systems seem reasonable, but just like the DSM itself need not be reified.

Here is some more information about RDoC for those interested.

http://www.nimh.nih.gov/research-funding/rdoc/nimh-research-domain-criteria-rdoc.shtml

Before transferring to research, I did psychotherapy for @ 15 yrs. While certainly imperfect, the DSM provided a common lexicon, and facilitated treatment planning. I'm cynical about this initiative, and wonder what might have been accomplished if the minds, that were occupied with creating this nosology, had been applied to treatment or research.