Psychiatric diagnoses are the map, not the territory. Many have overlapping symptoms, likely because the mechanisms are not all that well-understood. I would not go so far as to say that the statement "$person has ADHD" is semantically null, but it conveys far less useful information than "$person is likely to benefit from Ritalin".

Addendum: My instincts point towards rule consequentialism - as a finite limited-information machine, optimizing my adaptations seems to me like generally a good way to win the better outcomes.

Same here. I've been diagnosed (by multiple therapists/neuropsychologists/etc.) with ADD, while on multiple occasions, non-experts have thought I had (or might have had) Asperger's syndrome. Meanwhile, a few people with AS said that I seem fairly neurotypical to them. And I've been evaluated for it but not diagnosed. So I'm probably somewhere on the border (if there is a border).

I agree with Will Newsome on the significant deviation of LW/SIAI aspie-types from the "one/a few obsessive interests" criterion. Until I was 13 or so, I did have a fairly single-minded focus on computers (programming and such), but my range of interests exploded around that time. And thanks to my ADD, I don't tend to stay interested in any one thing long enough to get much done on anything...