I have a question about the GFDL and audio recordings of GFDL text (for accessibility, podcasts, etc): how does one comply with the GFDL in an audio version of a text?
Does it really require the entire license also read out?
This question was raised during the FDL drafting process (but I don't have a reference, sorry), and never satisfactorily addressed. It's yet another area where the FDL effectively makes a work non-free for many uses, and another reason to recommend against its use.
It is another reason to not recommend it for non-documents; a audio recording isn't a document.
My advice would be to convince the copyright holder to re-license the work under something more sane, like the GPL v2. That, at least, doesn't require the license terms to be included in the redistributed work.
That is a even worse suggestion. How will you distribute the written offer? Because neither 3(a) or 3(c) apply (audio is not source code, or object code).
The GPL was designed for software, audio isn't software, a suitable license free audio license would be be a verbatim only, or a all permissive one.