Interesting ideas, but when it gets into utopian promises about being able to find all versions of a file ever on the web, it reminds me of the talk around RDF[1] and FOAF[2] in the late 90s/early 00s. Not to mention pre-Web hypertext theory, PGP, and any number of other things. All these things can be very useful patterns in tightly restricted, highly curated environments with specific use cases (whether or not they are part of the larger web). But they aren't panaceas and they don't address the underlying sources of the problems they are trying to work around, namely that all this unorganized junk on the web they aim to take control of is created by people and companies who are one or more of: lazy, sloppy, confused, conflicted, error-prone, avaricious, malicious, obnoxious, cruel, naive, desperate, or ignorant.
[0]: https://joearms.github.io/2015/03/12/The_web_of_names.html