Please edit, it's a draft and I think it's a problem many of us know from a long time.

The idea is that, if we had a system for MediaWiki to know what a book is,
we could
* better metadata management system, thus a better integration with Wikidata
* we could develop better tools, for example create automatic Indexes and navigation templates, which would maybe enable us to "import" EPUBs directly on Wikisource
* have a better workflow, making it easier to new users to use us
* have even better statistics and analytics: which are the *books* (not the pages!) which are read o WIkisource?

etc.

IMHO, the Community Wishlist is a great opportunity to push for a systemic development of Wikisource: I'm convinced we don't need little tools here and there, but a systematic overview of what Wikisource framework, workflows and software.
I'm open to discuss the merging of different proposals in the Wishlist.

Aubrey