Nope, we're just learning to recognize it better, destigmatizing it so more people seek out a diagnosis in the first place, recognizing it's a spectrum and that people with a more subtle expression of autism can still be diagnosed, etc, etc, etc.

I know someone in their 50s who was diagnosed recently. They've been who they are their whole life. I know a lot of adults who have similar personality traits that haven't been diagnosed - I don't know how many would be, if they were to speak to a professional, but it's not 0. (I think there's maybe a 25% chance I could get an autism diagnosis, were I to talk to my doctor about it. People have encouraged me to, but I personally don't see how that information would be useful to me.)

These adults are the "missing" children from your statistics.

I think it’s at least relevant to note that a lot of things relating to autism was completely redefined in DSM-V. DSM-IV had many different diagnosis such as classic autism, autism spectrum disorder, aspergers and PDD-NOS (Pervasive Developmental Disorder - Not Otherwise Specified). All of those was merged into a single diagnosis titled ”Autism Spectrum Disorder”, where the criterias are communication difficulties and stereotypical behavior. My understanding is that this was mostly due to poor diagnosis stability with the prior set of diagnosis. It seems at least plausible that this general simplification of diagnosis criteria has contributed to an increase in the number of diagnosis. (It’s also worth remembering that any comparison over time has to bundle all of the previously distinct diagnosis to come up with an apples-to-apples comparison.)

Do you have any sources that back up increasing prevalism over better recognition and broader diagnostic criteria? Afaik almost everyone chalks it up to those 2.