Several reasons. The purpose of archival scanning is to preserve not just the content but, as much as possible, the physical details of the artifact itself. Blank pages can sometimes, when processed correctly, reveal content that has faded or been erased, particularly in very old books. It can be of interest to know exactly how a book was bound and where the pages lay relative to each other, and in a book like this one that lacks page numbers, that information is lost if blank pages are omitted.
Basically, best practice is to always scan absolutely everything, because a scanner tech is not qualified to judge what may be of interest to researchers. Hell, the researchers themselves may not be.
(That said, whoever prepares an archive for public presentation has a duty to make it reasonably accessible, which could have been done better in this case.)
