I started bemoaning the classification (aka ontology, nosology) of neuropsychiatric disorders about 8 years ago. I'm not the only one. One of the things I liked about Greene's Explosive Child book is that he is clearly unimpressed with the DSM IV nosology.
We're due for another DSM edition, but I doubt that will be any better.
The good news is that in the last 8 years it's become clear to every researcher that all of the common neurospychiatric conditions, from "ADHD" to "ODD" to "Autism" to "Aspergers" to "Bipolar disorder" to "Schizophrenia" are very rough categorizations of thousands of different "phenotypes" (where a phenotype is the end-result of the interaction between genes and environment) that are themselves dynamic over the lifetime of the brain. (Even after adolescence, we see major changes in schizophrenic symptoms between 20 and 50.)
Over the last 3 years we've seen that many different combinations of diverse gene variants, combinations, "malfunctions" and prenatal genetic express modification can produce superficially similar clinical presentations that we squeeze into the garbage bins of "mental retardation", schizophrenia, and "autism spectrum disorder". Most surprisingly, many brains with extraordinary genetic disorders appear normal.
This classification problem isn't simply an annoyance for researchers and industrial ontologists. It has important legal, educational, financial and, yes, clinical implications. The legal, educational and financial implications are large but outside the scope of this post. Suffice to say there is a reason that the diagnosis of "autism" has exploded while the diagnosis of "mental retardation/DCD" has shrunk (clinically speaking both diagnoses are about equally useless).
The clinical implications are what matter to most of us. Autism is a fine diagnosis for getting hugely beneficial school and family services, but if it leads to pure therapeutic choices or misguided interventions then it's harmful. If we lump too many conditions into one bucket, we risk choosing the wrong interventions because they don't match the bucket.
I'm hopeful that over the next decade we'll see a revolution in thinking about neuropsychiatric disorders, and the evolution and diversity of mind. We'll become more empirical about what works and what doesn't, and recognize that the brain at 8, 14, 17, 20, 30 and 40 may be very different. We'll always need classifications, but they may be more akin to "Ritalin responsive" than to "ADHD".
Progress has been slow, but it's coming.