If there were good (and fast!) js, c, python, etc libraries to fully support parsing org-mode to AST and HTML, only then would it be a viable solution for markup.
> there's no good way of converting <markup language> to html except by running <preferred converter>. So for each article, getting a html representation requires firing up <preferred converter>. That's insane.
But if you are content with a less featured variation of the markup language (but still more featureful than Markdown), you can get that for JavaScript too: https://xiaoxinghu.github.io/orgajs/syntax
not to mention the security hazards of running emacs with any user input from a website.
Second, "firing" up Emacs sounds odd to Org Mode's target audience. :) We already have Emacs open! (I typed this comment in Emacs).
I don't know how hard it is to make a nice looking website in Org Mode, but to make bland websites for academics (like this schmuck's page: https://www.matem.unam.mx/~omar/), I think Org Mode and org-publish are great. And probably it's not harder to make a nice website with org-publish than with anything else: the trick is being willing to learn CSS. :)
Of course the canonical format for pandoc is markdown.
Kulfon is my own project, so take it with a grain of salt. I've decided to focus on Org Mode as I also believe this format doesn't get the recognition it deserves. It doesn't mean Markdown is bad, au contraire, it is almost perfect. In Org Mode, for instance, I still cannot get used to the source code blocks. It is, however, good to have a choice.
The Org Mode support in Kulfon is not perfect yet, but good enough for most situations. It is based on a JavaScript plugin by Xiaoxing Hu called orgajs [2]. You can use it directly in your project.
One of the motivations for building a SSG with Org Mode was Org Brain [3]. I have a plenty of interconnected notes that I would like to publish on my website « as-is ». This would be rather difficult to do using Markdown format.
[1]: https://kulfon.org/ [2]: https://github.com/xiaoxinghu/orgajs [3]: https://github.com/Kungsgeten/org-brain
A partial snippet of the code that handles new posts:
https://gist.github.com/TeMPOraL/c401f12d50d4e1d4f94492e3d0f...
What's happening is that I take the HTML generated by Emacs Org Mode exporter, deserialize it to DOM, select a block corresponding to contents (to cut out superfluous stuff I don't need), and then run a bunch of tree-walkers that massages the DOM until it becomes something I can wrap with my site's scaffolding and dump into a file. It's really ugly, but gets the job done.
IIRC last time I tried this (long ago) there were a number of gaps in things like org-mode tables.
At the limiting case, the (common or uncommon) elisp extensions (e.g. org-babel) will likely never be processed by third-party tools.
The one problem I have with pandoc is that when using word export, it loses the table and figure numbers. Mind you, at least I only have to put those in outside emacs :)
To me that's got the same drawbacks as "not standardized". If not all of the existing tools support the standard then there are several de facto standards.
So far, I've not seen any major issues.