If, however, I convert the image to a non-proprietary format, I will lose some of the editability of the work. Which is more important, breadth of access or depth of access? (Hmmm. This parallels some current governmental debates on the arts here in the UK 🙂 ).
In my own work this can best be illustrated by the case of some images I made about ten years ago using Adobe Dimensions (a *vector* 3D package). I couldn’t have done the work any other way at that time, I can’t convert the work to another editable format now, so what do I do? I can release the all-but-unreadable editable format, or I can render the work as a 2D vector image which is editable as far as it goes, but does not have the full editability of the 3D version. I’m about to make some more of this work in Illustrator CS, again using its unique capabilities. As I said before, if one takes FSF-style Freedom as a guiding principle, I shouldn’t even be using Illustrator. But as an artist, I need various facilities it has until Inkscape catches up, facilities Inkscape can’t even render (full-strength masking for example), and I want to provide editable files containing those features.
In the absence of a guiding principle of FSF-Freedom, this is a dilemma. A producer will want the fully editable format, they will probably have the software. A consumer will want the accessible format. I provide various formats, but even with scripts to derive various formats for me from the original, this takes effort, and with (say) a movie, the space requirements would become a consideration.
I’m not arguing against editability or accessibility, far from it, they are vital. But I do want to illustrate that this may mean something (or some things) different for Creative Commons than for Free Software.