I'm happy to let you know that it.source community started a project to build one, or more, transcriptions "with scans" of Alighieri's Divina Commedia; presently it.source has only a "naked" version.
As you see, pages mix two different, indipendent texts, Alighieri's text ad Lana's comment, each one with its own, separate annotations. The author's aim was, obvioulsy, to allow an immediate comparison of the comment and original text. How can be solved the trouble of indipendent annotations, of reasonably simple use of pages tag, and of other issues, saving the author's aim too? Are there running examples of proofread works with a similar page structure, just to learn needed tricks from them ?
Alex