Adapt Tor Browser manual to use the new support pages

Meta issue to track how we can integrate the new support pages (https://gitlab.torproject.org/tpo/web/marble/support, https://support.staging.torproject.net/) into Tor Browser under about:manual.

@gus can you cc other relevant people from your end?

Some initial questions:

Will the old Tor Browser manual be available for the rest of the 15.0 cycle? Will it be updated as well?
As far as I can tell from https://gitlab.torproject.org/tpo/web/marble/support, the new support pages uses Lektor. Is that correct?
Can we get a general overview of how the contents.lr files are converted into HTML? In particular, how is localisation handled?
Is there some comprehensive documentation for the .lr format? Any tools for extracting standardised data? E.g. a python module? Or would it be simple enough for us to write our own parser?
In principal, we could write our own top-level HTML in the lektor format, and bundle the generated HTML into the Tor Browser build for each locale. Would this be feasible? However, there is a risk of a supply-side attack or bug in Lektor causing problems in Tor Browser. We could sanitise it to remove all CSS and javascript. But maybe it would be better to simply parse the lektor files and convert them into a JSON or something, and then construct our own HTML, CSS, etc on the Tor Browser end.
I could not find any accessibility documentation for Lektor. Is there some guarantee that the HTML it produces is actually accessible? If it uses standard HTML and sticks to the HTML semantic specifications, then it should be mostly ok. But if it is using non-standard widgets via some <div> magic, then I'm more concerned.

Side note: I had a quick look at www.getlektor.com: I've never seen documentation pages with adverts and comments. lol.