I’m so late in replying that I doubt anyone will read this but thought I’d add my 2 cents.
I see no discrepancy between Barry and Alan’s statement about bit depth. I scan at 16/48 bits and print at 8/24. The reason is that our scanners have an amazing ability to delineate the most subtle tonal variances and 16/48 allows us to stretch the ‘histogram’ and avoid any sort of posterization. However, like Barry I can’t see any difference when printing at excessively high bit depths…you can do it…you just can’t see the difference. I don’t even think it’s a printer hardware limitation but more of a human limitation when perceiving the constrained dynamic range of a printed surface.
So I scan into PS, crop, straighten if necessary and maybe a slight tweak in tonal adjust and save in the full 48 bit depth, thus keeping a pretty much pristine scan. Then I get to work and do whatever really needs to be done in preparation for print, then save again. Often lots of times. I’m a big believer in redundant saves.