Yep, you're right. That's why I called it an "extreme example" .

You can still see the surface texture about a third of the way across, and that's with spacing off the glass of about a quarter of an inch. You can do the trigonometry, but I maintain you couldn't see the loss of sharpness with a spacing of 5 or ten mills, a typical plastic page thickness.

Not trying to argue that reference scans should be made through a plastic page because there are other considerations, like dirty or abraded pages. And as my example shows, you should reall clean the scanner glass before scanning a chip .