You're right that diagnosis numbers alone can't tell us how much of the increase reflects better identification versus an actual increase in prevalence.
But improved identification isn't hypothetical. We know diagnostic practices have changed, and people who went undiagnosed for decades are finally being recognized. Someone diagnosed with autism at 40 didn't suddenly become autistic at 40.
That's the progress I'm talking about. More people getting answers and support is a good thing, regardless of whether other factors are also contributing to the increase.