The irony here is that Youtube videos from ten years ago are still alive and well. As Youtube makes a much better places for publishing and archiving content than the rest of the Web. With Youtube you don't have to worry about URLs changing or domain names expiring or anything like that. You just publish your video once, get a unique video-id and don't have to worry about anything else. Google's monopoly worked here in our favor for once and they have been reasonably good in not breaking old content (not perfectly, as video annotation got crippled pretty badly).
I think that's where the rest of the Web fell short. The Web has no concept of "publishing". There is no ISBN when you write a blog, no library were you could look up that ISBN. It's all just a file on a server or an entry in a database, that will get mangled and lost in the coming years. Worse yet, the article itself isn't even accessible from the Web, it's mixed together with a user-interface, ads and other stuff or spread across multiple pages. All this makes it quite tricky to keep old content readable and archived for the future.
This also leads to a weird situation that a lot of publications are still avoiding the Web after 20+ years. They publish as ePub or PDFs instead, which you can somewhat access from the Web, but really aren't well integrated into it. But it's by far the easiest way to ensure that a text document published today will remain readable a decade down the line.