Video games disprove a lot of UI/UX "wisdom" by demonstrating that people actually do learn - if they care. It's often the context that makes people not care. Do they get any personal value from caring? Can they get away with not caring and still get the same (or same enough) results? For video games, the answers are, in order, "yes" and "no", and so players easily learn to operate absurdly complex interfaces, and to do it at speed.

No, all you need to disprove this is to look at the achievement rates for beating a game. When barely 1/4 of people finish the game, which ideally is what a developer wants, the idea that players muddle through is a bit much.

I think I've finished like 5% of the games I played with. Even counting ones on which I spent more than 10 hours, I don't think it would be more than 10%. Hell, I'm 241 hours into Stellaris and haven't even won one game yet; my Kerbal Space Program hour count is four digit and you can't even win that game :).

Point being, people don't finish games because they get their fill of enjoyment before then. Many games I know start to feel like a chore after a time; some early on, others near the end. Unless there's a story you desperately want to see resolved - and many games these days ship without one - there's little reason to play a game once you mastered all the different mechanics it offered and seen most of the variety it featured. And then there's multiplayer, which wouldn't even count under this metric.

I propose a different one: people who play a game with nontrivial UI for more than 2 hours are evidence in favor. If the UI was a problem, they'd drop out sooner. After all, unlike software used at work, nobody forces you to play a game.