I'm baffled as well. I think it's fairly understood that any white pixel is always understood to be probably overexposed.

Astronomy often also deals with wide ranges of pixel exposure such that it can be hard to tell the difference between overexposed and "brighter than the max on the color scale". Better to be extra clear and remove all doubt.

Additional bits of information? You cannot see what shape is that area if it is surrounded with a light yellow and you will miss some info if you'll do it after JPEG conversion.

> Just to make it extra clear that the pixel data exceeding overexposure threshold is not mistaken for valid data?

Saturated pixels are not necessarily "invalid" the way cosmic ray bit flips and stuck pixels are, but setting them to zero does make the other bright but not saturated pixels much easier to identify.

Does the precise line dividing "bright but just barely not saturated" from "bright and saturated" actually matter?

Yes, because a saturated pixel might be 101% full scale, but it might also be a hundred million percent and you have no way to know. The same reason that a camera might substitute zebra stripes or a bright colour for a blown highlight in a preview as opposed to just making it white.

Though the actual value in the data isn't black (0,0,0), it'll be some obviously special value that won't otherwise happen, because black is a valid value.

Because that’s what the person who was dealing with the data did