Here's what I can think of off the top of my head:

• Page sizes, DPI, color depth, and a check sum are stored in the database
• Pixel locations might be used by QF for things like Zone OCR
• Annotations are stored by pixel location
• Record retention laws may specify which compression types are acceptable
• OCR'd text is indexed against pixel location

There are probably other things that rely on the integrity of the image. Pixel locations may not be a big deal, as long as the DPI and image size aren't changed, but the other issues make that moot. IMO, the idea of any modifications to the volumes that aren't done by Laserfiche is frightening. One of the selling points of a system like Laserfiche is that you trust that nothing will change without your permission.

If somebody was really concerned about the size of their volumes, you could always look at using the SDK. The Laserfiche.Imaging namespace would allow you to re-compress pages in a more "approved" and safer manner. You'd have to do some testing to see how stuff like annotations would be affected, although as long as the DPI doesn't change, it should be fine.