I don't think anyone would deny Lynch had his own, distinct Lynch-ian stye. A lot of people probably can't describe it clearly in words but they'll recognize when they see it. So I think it's plausible LLMs "getting the gist" of the style and asking them to generate something "Lynch-ian" might just work.
I propose a Turing test for it. Show a bunch of people unfamiliar with most of his work some of his works mixed in with some generated, and see if they could tell which is which. It will fail today, but I think in 10 years we'll get there, even if not sooner.
I'd say it is, although I can understand it is anti-thetical to what his fans get out of his films. Objectively though, I think you can certainly make a list of defining traits and show a pattern.
> David Lynch didn't pull out a notepad and go, "Okay, time to write a Lynch film." He just decided to make a film and made it the way that made the most sense to him.
I'm well aware, but it doesn't mean there isn't consistency or a pattern in his results. Plenty of filmmakers have a particular style and have elements that show up in anything they make allowing people to recognize their influence.
> At best (and this is seriously stretching the definition of "not just wild speculation") I think it's more likely you get something that exhibits formal qualities that remind someone of some of David Lynch's works, probably a fairly incoherent and incongruous mash of tropes, like a particularly bad episode of Family Guy.
So to discuss this further in detail, I think you'd need to at least try and define the ineffable qualities you think an AI wouldn't be able to recognize (keeping in mind eventually AI will be able to take things like your comment history into account to help it 'understand') to help it create something that could have been made by the man himself.