Not a crazy thought, in a more general concept.
First off incidence.
Depends on who you believe and how you define and how you case find. Problem is that autism is not a clear either/or entity. It is part of a spectrum of disorders - the autistic spectrum disorders. Where you draw the line matters. More on that later. One in 500 is an upper estimate, one in 2000 more standard.
Genetics:
Large genetic component, probably greater than 90%. But not just one gene. Maybe three, maybe as many as fifteen different interacting genes. We know that sibs of autistic children have a greater than 50X greater chance of having autism than the general population (3% to 6%), that fraternal twins have no increased rate than other sibs but that identical twins have a 60% concordance rate. There is a rapid fall off for more distant relations. These facts are what provides the basis for those estimates. No single gene is a cause - autistism is a common end phenotype of multiple etiologies in a complex system. Future research must address why this is such an “attractor basin” for perturbations of the systems of neural networks that is the human brain.
Broad autistic phenotype:
Relatives of autistic children, who do not have the label of autism themseves, are statistically more likely to share features of a broad autistic phenotype: speech/language problems; impulsive, shy, aloof, eccentric. They are more likely to include engineers and others who make their living in math and programming related feilds.
Is there more autism that in years past?
Certainly more diagnosed. Could be a real increase, maybe not. Some can be increased case finding and a broader defintion. We docs look for it now because we have become convinced that early identification can lead to better remediation. Give me a good intervention to offer and I’ll do a good job finding who needs it. Consequently kids are labelled as having autism today at age two that never would have been so labelled a decade ago. Kids with the label are elgible for more services. Services that we are convinced make a difference.
Assuming a true increase, is there any reasonableness to the implied hypothesis of the op?
Yes. In today’s tech-happy world many socially awkward eccentric boys and girls (who would have been shy social outcasts with poor economic prospects a few decades ago) are meeting each other on the job and elsewhere and are considered desirable mates on the basis of economic success and shared interests. These members of a group that includes those who may have that broad autistic phenotype, each with one or the other of those independently interacting genes, are mating with each other in numbers previously unheard of. Match.com wouldn’t be the cause, but it would be one example of how such occurs more commonly in today’s world than in times past.
I don’t know if there is more autism or not, but if there is, then this hypothesis is as attractive of an explanation as any other out there. Better than many others that get more press.