As this is a topic near and dear to my heart, I've looked into the neurobiology and genetics a bit, and I think that the actual science is showing us a much different perspective than the proposed DSM-V revision would suggest. It's pretty clear that Asperger's Syndrome, "classic" autism, NVLD, PDD-NOS and Childhood Disintegrative Disorder, far from being all the same thing, are associated with different groups of genetic defects that have different consequences on neurological development (with some of them being related to auto-immunity against microglia, others affecting the dopamine system, some affecting cerebellar development, etc.,) and that they have very different clinical presentations, although they share some common features. Some of the "spectrum" effect is due to individual interaction with environmental factors, and some is due to some individuals having multiple gene polymorphisms.
Diagnosis means "to know between", and diagnostic categories should highlight, not obscure, clinically meaningful differences, so that finding the appropriate diagnosis automatically directs clinicians toward appropriate treatment strategies. The appropriate interventions for a highly verbal, uncoordinated individual with Asperger's who, as is typical with Asperger's, has difficulties with visual processing and visual-motor integration are very, very different from those that would be appropriate for a non-verbal individual with classic autism who has visual-spatial strengths, and lumping the two of them together does not make sense either from a clinical or neuro-biological perspective.
