I've encountered a few from Archive.org that display mostly blank pages on my Kindle K3 although they display properly in Acrobat Reader on my PC. All of these are created with separate layers including one for a background image of old paper and a transparent layer above it with the scanned text. I've "fixed" them by printing from Acrobat Reader/PC to a PDF virtual printer to create a "flattened" layerless PDF. The resulting PDF displays fine on my K3.
Printing this way can be slow as the necessary rendering and image merging is processor intensive (probably related to why they fail to display on low resource ereaders). Some took hours. The virtual PDF printer I use is one I've used for several years,
PrimoPDF.