You may want to have a look at “local copies” in the manual.
Most of those flags are probably there for debug purposes. From the code, a number are also stored with the image data, and just “repacked for display”. The ones that are directly useful are also visible elsewhere (e.g. ratings, local copies, raw file) or not useful (like the loader information).
But if you are evaluating darktable to see if it fits your needs, why are you even worrying about this? I’ve been using darktable for years, and never needed any of those flags.
Same thing with the metadata items provided: they are sufficient for my needs.
I did look at the IPTC system, and found it too cumbersome for my use case: there are just too many to fill out. Then again, I’m not a press agency handling 1000s of images from multiple sources all over the world, with associated rights, embargos etc..
And of course, each metadata field requires a column in the database. And all that has to be maintained.
As for documentation: it’s hard enough to keep the basic documentation up-to-date… A “documentation” category wouldn’t change that (and why a global category where documentation is specific to one program?).
More official documention would shift the burden from the many (e.g. this forum) to the few (the doc writers). In addition, “how to” (semi-)official documention could create “the one way” to do something.
One of the characteristics of darktable is that there are often several ways to do something, which each have there own trade-offs (speed, scale, risk of artifacts in some situations, …). That shows up clearly in questions here.
What also shows up is that some do not read documentation, but just start using the program and ask questions when they hit a snag.