I think what Amanda Baggs writes here might be an eye opener:
http://ballastexistenz.autistics.org/?p=79
The problem is that when you disagree with something in a single post, you tend to attack the whole post, which is easily perceived as a personal attack. And when you say "no evidence" it could actually be perceived as ignorant.
You need to learn the art of gentle criticism.
Actually, "autistic psychopathy" never meant psychopaths.
"Psychopathy" in that context meant "psychological pathology". It did not mean sociopaths, psychopaths, whatever.
I'm with EvilZakkie on this one. There's nothing that suggests autistics are any more likely to have children, or even as likely to procreate, as NTs.
A couple of points. First:
In a discussion on this board earlier, I asked how it could be that there's an almost constant population of homosexuals in society despite them (generally) not reproducing (well, at least not until modern technology with artificial insemination and whatnot came a long). I had naively ignored that a trait, however genetic it is, need not be apparent in any given phenotype. Genes can lay dormant for many generations; one common example is a gene that causes some deadly disease under special conditions but under other conditions protects from malaria.
One doesn't have to be overly social to have children, but it certainly helps. I cannot believe you think that a group of people characterized by impaired social understanding and lower social drives have as many or more children than do everyone else. It makes no sense at all. To spread your genes, you probably need to be in a relationship, at least nowadays, when contraception and abortions are easy to come by. To do that, you need to be possess some social skills. And especially when you consider the lower-functioning end on the spectrum (by lower functioning, I mean simply unable to function in daily life without substantial help)--how in the world do you suppose they can keep up with everyone else when it comes to spreading genes?
Even in Silicon Valley I doubt people on the spectrum procreate more than others, and that area is commonly thought to have an extremely high density of autistics compared to everywhere else.
Second, you mustn't so easily cross the is-ought gap. The Scottish philosopher David Hume noticed some centuries ago that many writers, when it comes to morality, make sudden, illogical jumps from what is, i.e., how matters are, to how things ought to be. This led him to issue a warning: be very careful when you try to go from is to ought. Some think it can't be done; I'm inclined to agree.
But anyway, this is what you do when you go from genetic traits to human value. That way lies Social Darwinism, the belief that society ought to be structured after natural selection: the survival of the fittest. Screw autistics, poor, and other people somehow disabled in the race for survival and glory: the best fit are the ones that ought to survive, and only them, for the betterment of society and the human race. I strongly believe this isn't what you think, in fact, I believe you think the opposite, but this is what can result from carelessly jumping across the aforementioned gap.
So, I'll agree with EvilZakkie: autism advocacy is about human rights, not about science or evolution.
Then again we can either not assume that autistics are unlikely to get children.

In the last hundred years fewer children are born per person at least in the industrialized part of the world. There is probably a big number of neurotypicals who never get children either.
Question is, what is the ratio between neurotypical and autistic reproduction.
Ofcourse what you said sounds logical, but I question whether it is a significant factor.