As per usual Ken is being pretty disingenuous here, even ignoring the hyperbole about needing a 175MP digital sensor to 'capture the same information'.
Velvia 50 might resolve 160 lines/mm (I'm not sure if this is line pairs/mm or lines/mm) but that's at a contrast ratio of 1000:1, for a contrast ratio of 1.6:1 (more typical for normal shooting conditions) you're down to 80 lines/mm which is 22.1 MP, and that's for just the film stock under optimal conditions. When you factor in real life shooting conditions like the lens MTF charts, the aperture you were shooting at, whether you were shooting on a stable tripod, dust, grain, noise, etc, that number only goes down.
Conventional wisdom has been that you can get good results scanning colour film at 6-12MP, and this tracks with the results people got when they started comparing the first DSLRs to 35mm film - something like the full frame 11MP Canon EOS-1DS from 20 years ago with a mere 25 line pairs/mm (including losses from the Bayer and AA filters) compares favourably(*) to medium format film, with 6-8MP DX format DSLRs from the same era comparing favourably to 35mm. B&W film can be scanned at a higher resolution but still, 87MP is a theoretical maximum not a practical one.
I took a photography class in middle school. We used 35 mm black and white film. I scanned all the negatives in with a decent Nikon film scanner. I can't remember why I settled on scanning at 2MP, but you could already see the grain: https://i.imgur.com/dcFksGp.jpeg
If you still have them, and a digital camera with a macro lens, you'll be surprised how much better result (actual resolved resolution versus MP "resolution") you'll get with than using a scanner.
If I had to guess as why you stopped at 2MP, it was probably a mixture of diminishing returns in terms of extra resolving power and size. Idk when you took them, but I imagine size was also a consideration. A 2MP uncompressed tiff (which I imagine is what the scanner gave you) is ~5MB. Now, you could drop that down to ~1.6 if you could save them as single channel rather than RGB.
This is also where the optimal conditions come in (motion blur, camera shake, exposure, and how it was developed).
It was! I scanned them around 2009, so it wasn't size, it was that I wasn't getting more useful resolution. The scanner was a Nikon Coolscan V ED. I might still have some prints I can look at to compare, but I'm not sure how to do the same with a negative.
That particular one offends me the most - we went from 0.3 megapixels to 300 and it's still fucking blurry