If it helps, current accessibility standards in epub3 are typically looking for something like the following as a pagebreak:
Code:
<span id="page_N_" epub:type="pagebreak" role="doc-pagebreak" aria-label="LABEL_FOR_PAGE _N_" />
Conventions can change but since a link to this id will bring you to exactly this point, the N in this case is typically the start of page N (or the end of page N-1).
Once you add these spans to match the pdf pagebreaks, you can use Python Function replace to actually build a list of filenames, and page ids that can then be easily edited to create a PageList for the nav.
Alternatively, you could use a word count to space out these pagebreaks but then matching them to a pdf page break will not be easy.