I have a gut feeling that one cannot really make any fundamental rulings in this film/digital comparison that has sprung into life here.
Take Velvia 35mm, for example: I have just spent time scanning on a CanoScan FS-4000US in order to try to turn a not particularly wonderful colour image of some old olive trees, shot through a piece of non-optical quality glass that was lightly smeared with vaseline, into a black and white version that seems to offer greater scope. The scanner claims 4000dpi resolution - why should it lie? - and the main surprise I have with the b/w version is that foliage has taken on what seems a bit like an infra-red look. Totally a surprise. Further, since the original was shot as it was to make it look fuzzy, I have not sharpened the b/w either, mainly because on looking at a test with the unsharp mask employed, it lost much of the streaky effect - pointless. Thatīs one kind of result with fine-grain colour transparency.
Using the same scanner on Kodachrome Pro 64, people/skin shots, the detail from the scans is just amazing, far better than I ever saw from the same series when printed four colour litho for the original calendars. Water droplets on the skin simply stand out as if they were real. I scan using none of the dust removal devices or anything at all that I fear could interfere with the basic detail thatīs there, and these shots too are now printed b/w.
So the point I might have made is that two different fine-grain films can look totally different when scanned, depending on many inputs other than just the scanning.
On the other hand, scanning fast 35mm b/w film can produce prints that look a little more crude, and slower ones can look pretty close to wet printing but with some micro control that would have been pd difficult to do in the wet! I have never scanned any 120, b/w nor colour, for the simple reason that I havenīt that size a scanner. Would have loved to have had the opportunity, though; Santa?
What about digital, then? I have only a modest D200, but the impression I get is that there is a gain in sharpness (using the same lenses) that could well have a lot to do with flatness of the sensor plane as compared with film. Trouble is, to come to any sure conclusion becomes difficult if only because of the unsharp mask step in the production of a print. With film you always had a loupe if you wanted to check edges against centre of a slide (and even with a loupe you could see the differences); with digital you donīt have that choice and looking at 100% doesnīt feel the same either. Such a subjective business.