[RcouF1uZ4gsC]

That is nice in theory, but in practice stuff like archive.org are vital. If you see a document you want to refer to later, you need to archive it, either in a personal archive or via archive.org.

There are too many moving parts to trust that even domain names will be the same. See geocities and tumblr for recent example. If you want a document, you should have archived it.

[JadeNB]

The article isn't arguing that URIs don't change; it's arguing that they shouldn't. (The part involving judgement is elsewhere in the title—the word 'Cool'—so it can certainly seem like an assertion of fact rather than of value at a glance.) It thus seems to me that the response "in practice, URIs do change" doesn't undermine that point; your discussion of the need for some solution to the problem rather supports their point—if URIs didn't change, then there wouldn't be a problem to be solved.

(Or maybe your point was deeper, that one not only can't trust that the resource location won't change but even that the resource itself will still be available somewhere? That is true, too! But saying that archive.org is the solution is just making one massively centralised point of failure. That doesn't mean that we shouldn't have or use archive.org, but that we should regard it as just the best solution we have now rather than the best solution, full stop.)