The article's main insight: "URLs based on hierarchical names are actually the URLs of search results rather than the URLs of the entities in those search results".
In the most technical sense both are searches encoded in to a URI form. The search for the (hopefully) GUID just happens to be for a specific mechanical object, while the other is describing the taxonomic categorization of what a matching item would look like.
Though their "/search?kind=book&title=moby-dick&shelf=american-literature" example is fundamentally different in that all filters (being URL query parameters) are optional and can be arbitrarily combined.
I didn't quite understand the point of the hierarchal "search URL" when you have the /search one implemented, and they go on to say you could implement both if you have the time and energy.
The Internet Archive WayBack machine kind of has an optional filter in a traditional URL scheme - you can replace the date in a WayBack machine URL with an asterisk as a wildcard and you'll get either the only entry it can find or a list of dates
Other links:
https://www.demilked.com/birds-flight-paths-dennis-hlynsky/
https://twitter.com/Rainmaker1973/status/1426785726239166465