It has been their business model for a long time! Give you the base figure, and then build clothing and accessories that work with the base models and sell for monies. A lot of their users prefer the adult type content, though, so going from there to meat-and-potatoes gaming is a stylistic challenge 
Also, those assets don’t come with redistribution rights, even for baked-in assets in games.
Some things they’ve done are pretty smart! Certain kinds of clothing or morphs are expressed as deltas off of the base mesh. You don’t build a camisole by adding new geometry; you build a morph target that shifts the appropriate vertices a little bit. Then, if someone else applies a “chubby” body modifier, that’s OK; the deltas will largely still work out alright.
Same thing for facial structure: you don’t build a new topology for the “Thai face” versus “Han face” versus “Basque face,” instead you move the existing vertices around to match.
The draw-back with this is that you need sufficient vertices in the underlying mesh to actually allow this, but with modern game characters being in the 30k+ vert range, that’s starting to become reasonable.
(I did this with 2k vertex avatars back in the early 2000s, and it was a bit more of a challenge there 
)
What I’m talking about is a standard for making mix-and-match paper doll assets. You can choose to follow it, and users of the paper doll system will prefer your assets, or you can choose not to, in which case nothing is different from what it is right now.
The discussion of “Art” vs “art” vs “entertainment” vs “product” is of course age old. That being said, if someone expects a pay check (or a marketplace purchase) then they should probably produce what those with the money (customers, project funders, etc) need. Or, you know, move to Humboldt county and join a hippie commune; that works, too!
Anyway – I’m not quite proposing the All Points Bulletin character builder, here, but any game with a large and composable set of assets have built those assets to a standard. Any RPG that changes your character when you put on different armor, has these kinds of rules. Any online 3D space (be it Mees or IMVU.com or Roblox or whatever) will have these rules, although only some of them support user-generated content. Plenty of artists exist who have experience building things like this, so clearly it can be done, and that “one item that doesn’t quite work” is TOTALLY FINE.
Currently, it’s not possible for me to buy, on the marketplace, a tunic, a leather jacket, an overcoat, a chain mail, and a plate mail, and swap them out on a character. There are plenty of individual characters that have one of those items, but there’s no standard set of guidelines or validation tools, so it’s not even POSSIBLE to build interchangeable content, even if you wanted to.
Having a standard is necessary for “paper dolls” to be possible. Having that standard strike the right balance between “it actually works for many real-life cases” versus “it’s an absolute pain in the ■■■ to develop for even for the best meaning artist” is hard, and can make or break the idea. Having enough initial pieces available to bootstrap the demand among creators is probably crucial – there are plenty of “character builders” on the marketplace, but most of them have way too little geometry, and none of them actually come with a design bible for how to actually build more content to match.
And I think that critical mass is the important bit. If someone provided a base character, and sufficient additional pieces to make for a minimal cohesive whole, and the design files, bible, and validation tools (!) to build more, it wouldn’t be a “set of rules published by a nobody,” but an actual, working, foundation. Some creators might find that of interest, and certainly a number of customer on the Marketplace, would!