Part of this thread seems to have morphed a bit into comparing film (especially transparency film) with CMOS sensors. (My OP excluded transparency film, for several reasons, but one was because I knew some people feel that scanned transparency film has exceptional qualities).
Comparing analogue film quality to digital CMOS image quality, even from a simplistic objective resolution perspective (let alone overall image quality), is not trivial. Technically, the Shannon sampling theorem should help (roughly stated, the pixel sampling should be at least twice that of the finest detail being recorded), but it's that "at least" part ... in the presence of noise/grain ... that makes the comparison tricky ... many practioners, based on their experience (not theory) would say it should be 3X-4X, not twice. I spent quite a lot of time in my past career comparing analogue and digital images, and it's not straightforward even considering only purely objective measures (which are just part of the story of course).
In my opinion, it is the appearance of the final image ... in the eyes of the person who made it ... that is the only thing that matters, and inevitably that means two people looking at two images, one taken with film and the other with a CMOS detector, will disagree as to which one is "better".
But my OP was not really concerned with technical image quality, but more with the rationale of shooting colour negative film in an old camera when you are then going to scan the film; my question was, why not just go digital all the way?
I'm not taking sides and by the way, there is no correct answer to my original question! All opinions have validity.