Yes, I've seen the speculations before, but they did truly seem anecdotal.
In this case, if the reporter is telling the truth and the doctor is telling the truth, and there doesn't seem to be any good reason to doubt it, then you have a sample of 30,000 children, not one of whom has autism.
If true, that is not within the margin of error and not anecdotal. There's no way that some other factor, such as an interest in home schooling, could account for it.
Of course it's always possible that somebody is simply wrong, or lying, or that all the other doctors in this medical group are lying to the boss. But this is still the most persuasive account I have read yet.