> ... and nearly no-one re-writes an academic paper afterwards for it to be more understandable.

Survey papers would seem to fill that role. They are a useful step for information moving from being on the research frontier to being in a textbook.