Depending on who's doing the advocating, plain text doesn't inhibit images or any other media, as any site with media written in markdown or reStructuredText shows. Lots of GitHub READMEs are written in Markdown and include images, for example.
There are a number of image formats that are commonly supported, as well as video and audio. Once you've decided to go beyond face to face speech in the same language, you have to make compromises. Just try to minimize those, and try to stay within conflicts that have compromises, like line endings.
I think starting from "authorship in plaintext," along with sharing plaintext files, would move things pretty far forward. (I guess there's some irony in there.)
There are a number of image formats that are commonly supported, as well as video and audio. Once you've decided to go beyond face to face speech in the same language, you have to make compromises. Just try to minimize those, and try to stay within conflicts that have compromises, like line endings.
I think starting from "authorship in plaintext," along with sharing plaintext files, would move things pretty far forward. (I guess there's some irony in there.)