> There is not a future where iPadOS "wins" and MacOS is slayed in ritual combat, or something.
Well, no, it's not about the Operating Systems; it's about the UI paradigms themselves.
I believe that Apple is worried that the desktop WIMP UI itself is going to be disrupted and fade into irrelevance due to 1. "do everything on the web" devices like Chromebooks, and 2. VR/AR productivity-workflow paradigms that are yet to be formalized; and so they're trying to find a successor to the WIMP UI, one that will still be relevant in 2040, even if/when the WIMP UI dies.
The distinction between iPads and Macs isn't arbitrary, insofar as iPads don't require a keyboard, and Macs do. iPadOS (and iOS) apps have to be designed under the assumption that a keyboard is optional; and that really changes things about how an app can work. You can port apps designed for keyboardless tablets to macOS just fine (and as of the M1 you don't even have to, you can just install them); but fully-featured macOS apps can't just be thrown onto a keyboardless touchscreen. (And you can't say "well, you can't install 'keyboard required' apps if you don't have the keyboard", either; the OS has no idea if the user owns a keyboard but just hasn't bothered docking it.)
You are accidentally correct in the other direction, though — that there's no reason you couldn't run iPadOS as well as macOS on a theoretical "touchscreen Mac" (which would be a different thing than an iPad, precisely because the keyboard would be welded onto it, and so apps could guarantee its presence/require you to use it.) The reason that Macs don't have touchscreens, AFAICT, is because Apple wants to use the iPad — along with everyone who buys one, and every developer who signs on to develop apps for one — as an isolated laboratory to run this "successor to WIMP UI paradigm" experiment. They don't want to "dilute" that experiment by allowing those users and developers to get the advantages of the iPad from any of their other products — because then those users and developers wouldn't be incentivized to use the iPad, and thereby to give them the experimental data they need.
Consider: why didn't Facebook just merge Instagram and WhatsApp into features of Facebook Messenger? And why is it still considered a huge mistake that Twitter killed Vine after acquiring it? Because, like the iPad, these alternate experiences — despite being owned by the same old bigcorp that serves you the traditional experience — are both innovation laboratories, the learnings of which can be folded back into the regular app; and also hedged bets against the market failure of the old-school experience. Vine could have been TikTok if it had been allowed to grow for a few more years. What could the iPad's UI paradigm be if allowed to grow for a few more years?