Film emulsion/chemistry has a specific “fall-off” curve in the way that it saturates when highlights hit over-exposure. Digital highlights clip more abruptly, which - while probably more accurate in an absolute metrology sense - makes for some compromises when making images of scenes with a wide range of lighting. (not unlike the difference between digital and tube distortion in audio, effectively hard vs. soft-clipping). There’s probably a case to be made that these analog modes complement human sensory systems adaptive nature, (eyes and ears are constantly modulating their dynamic ranges). Lately differences in imaging are less stark with the advent of higher bit-depth image sensors, allowing greater exposure curve adjustments to be made in post. Relatedly Look Up Tables are all the rage for creating sophisticated color mappings similar to what used to be the purview of labs.
eta - … and specifically to the question of whether it makes sense to capture with film even if scanning will occur later, yes - because of the differences in the amount of control that one has at each of these stages. Capturing on film for later scanning is a big thing - especially with larger format stills photography (although in that case it’s as much about the optical system as the film). High quality scanning is inherently slower because it takes more (and more well controlled/balanced) light/exposure time than is convenient at initial capture.
