What works for the Linux community would be an unmitigated disaster for the IETF which is supposed to be trying to serve the users of the Internet.
The reason HTML email exists is simple: The users decided they wanted it. And when the IETF did not deliver an HTML email standard that worked, every vendor went off and invented their own riff on the theme. And so we have the worst of all worlds.
I very deliberately use what the general users of the Internet use because I want to understand the Internet as they do.
Now could we have done better? Well Nathaniel Bornstein attempted to deliver rich-text when he was working on MIME and got a lot of grief for the effort.
It is really not too late to do the job right. GMail has a massive footprint in the market place, the other major email client is Outlook and that is mostly subscription based. So if there was a coherent HTML-like email markup and two vendors support it, it would have a very good chance of success. That is not to say nobody else gets a say, just that if those vendors are on board, good things can happen.
One of the huge problems here is that HTML itself has been massively corrupted. It is no longer a document structure markup, it is a document presentation, active code, a floor wax and a desert topping. HTML/2.0 is a sensible markup for email, very little of what came after is.
What we need is more markup, not less. Fisking a post often comes to grief because the various annotations interact poorly and then there are the people who forward a 200K post with one comment right in the middle.
Oh and we are still suffering from the brain fart someone had 40 years ago when they decided that they would 'solve' the problem of the lines wrapping at the wrong place on their VT100 by wrapping all the lines in all the emails at 72 chars.
I am currently working on the GUI for my end to end encrypted social media, messaging etc. client Everything:
I really don't care what the markup is, but looking at existing practice there are three very clear sets of functionality that are widely observed:
1) Chat/Messaging, short posts with only line oriented markup, bold font, fixed font, etc. Examples, Signal, Slack, Discord, Matrix, etc.
2) Mail messaging, longer posts of letter length. These require paragraph level markup, headings, code fragments etc.
3) Full document structure. For this XML2RFC is actually pretty good albeit some of it somewhat IETF focused.
I am really not a fan of 10 stack deep email comment threads, I think that annotations on a text need to be seen as something different from the text itself.
So the rules for my annotation system are that everything starts with a primary text which is typically either a mail or a full document. Then people can annotate the text with the single line chat markup. And those annotations can in turn be annotated and the document and its (selected) annotations can be presented as a new document.
But if you want to make use of paragraph level markup, you have to do that by creating a new primary text which can quote the original and its annotations as appropriate.
