But also, it's a distinctly different answer for each page to build its own toolkit for the user (Hyperclay) vs TBL's read-write web. The user-agent ought, imo, afford standard tools that are going to work across web pages, that extend the user agency whatever site they are visiting.
https://www.w3.org/groups/wg/lws/
That would likely have some overlap.
If you were to have an accepted w3c proposal and working implementation in local browser forks, you could potentially chat with the browser teams to add the experimental feature first through a flag users would manually have to turn on, and then later potentially it could get integrated.
Isn't that basically Wikipedia? I can't imagine a much simpler system that could work at modern web scale.
Saying "we don't need this because Wikipedia already solved it" is kind of like saying in 1976: "Nobody needs the Apple II, we already have IBM mainframes that have solved every useful problem in computing much better."