It's not clear how distinctive a lot of mental health conditions are. For the most part we're just making up labels for groups of signs and symptoms, after all. To the extent the labels are "real" things mostly rests in their utility—I mean, that's sort of true of all labels for everything (what's a chair?) but these are far more fluid than most.

It could be that autism really is, exactly as we describe it and conceive of it, in some meaningful way defined by actual reality, a thing.

It could be that it's ten different things that aren't actually connected at all, but happen to look kinda similar. Some of which either are a variant of ADHD (or vice-versa, doesn't much matter) or just happen to include similar symptoms and behaviors that respond well to the same drugs and therapies we use to tread ADHD.

(now, to some degree we do have real tests we can do to pick up e.g. genetic markers of certain disorders, but these largely remain just another clue, not exactly solid proof, with some exceptions)

To illustrate: imagine we couldn't ever see the inside of a human body and just had to guess at what was going on when something went wrong. We'd probably have something we just called "bad kidney" that was actually several different problems, and we'd just throw drugs and other therapies at it until (often, but not always) some set of those relieved the symptoms. Meanwhile, sometimes it's a kidney stone, sometimes it's cancer, et c. And maybe we even have a whole step that's trying to figure out if it's "bad kidney" or "bad bladder" and sometimes we'd get that right, but sometimes wrong, but also some of the same medicines work for either (depending on the actual cause) so we might incorrectly diagnose "bad kidney" then accidentally correctly treat "bad bladder", and think we were right all along.