I don’t understand how even the preference for shooting on film makes any sense if you’re digitizing, heavily processing, adding entirely digital elements …
I get putting the final edit to film.
If you’re not doing all that additional garbage, then capturing on film, going digital for basic editing, then back to film might make sense.
The end result is not recreatable with digital assets. i.e.: It doesn't matter your budget, we cannot currently simulate the look of the original footage shot on 35 (let alone 70mm, forget about IMAX 80mm) by shooting digitally.
Also what you're referring to as 'garbage', is colour grading and composite work. It's not adulteration, it's an essential create part of the filmmaking work. Whether you're doing it exclusively as part of the development process of the physical film, or digitally in Nuke or Resolve etc.
Your inability to understand why people do things in a domain you're not familiar with is indicative of gaps in your knowledge, not the futility of the process. Lots of information about how and why all this works as it does online.
I think what makes film great is more the process. It forces slower more deliberate filmmaking and somehow that seems to show up in the product. Only celebrated/respected filmmakers have the influence to force production to allow them to shoot film so there’s a selection bias at play too.
That is not what Yedlin's tests show or made an effort to test! His famous first set of comparison tests are about resolution, specifically his opinions about the unimportance of 4K.
I deeply disagree with his conclusion, because he is downgrading a digital image to show it can match the resolution and some of the visual aspects of a 16 print. Which is like putting a filter over a high resolution monitor and then comparing it to SD TV.
But in either case, his work wasn't about whether digital could match the look of any film (it can't).
He has also never released his nuke workflow, although he's more recently productised it to some extent.
I cannot speak to film->digital->film but for audio recording if you did a similar thing like tape->digital->tape you would impart some real stuff to the recording versus just digital all the way through.
This is out of my area of expertise a bit, but digitization quality loss often depends on whether you quantize things, at what level of fidelity and also what kind of transformations you apply (stepped discrete or continuous for instance).
Consider how music production is often done on audio recorded at a higher sample rate, so when effects are applied artifacting is reduced for the eventual export sample rate. For video, one analogy is recording slow-motion shots at a higher framerate rather than slowing down regular footage.
I get putting the final edit to film.
If you’re not doing all that additional garbage, then capturing on film, going digital for basic editing, then back to film might make sense.