Scholarly publishing is an area where documents need to stick around for a long time, be able to be cited in a well-understood way, etc etc.
Crossref is an trade association of scholarly publishers, formed to solve exactly this problem. Part of the deal when a publisher joins Crossref is an obligation to keep the links up to date and a plan for what happens if the publisher suddenly vanishes into thin air.
Initiatives like CLOCKSS address the archiving question http://www.clockss.org/clockss/Home . The CLOCKSS homepage has a good overview.
And to the 'tracking stuff down' point, the DOI is resolvable (i.e. a link you can click) but it's also an ID that you can look up in a database. DOIs predate the web and may live longer than it. Having a single, 'official' ID (whatever it looks like) is better than searching by metadata.
As someone who works with scholarly data, having IDs for things is absolutely invaluable. There's a lot you can do with DOIs quickly and easily that would be prohibitively difficult if you had to use human-added metadata.
It's good to hear that scholars are getting together and trying to mitigate the volatility of URLs. I'll definitely need to read more on the subject to understand the pros and cons so thank you for your introduction in the article; many specialist blogs tend to only write for their specific audience. It seems like DOIs are an ample solution as long as they aren't a single point of failure.
Here's a short list of posts that will give some background:
A DOI is a Digital Object Identifier.
A digital object is something like a scholarly article. People need to be able to find it by URL, cite it, and refer to it with an ID. The URL is no good because it could change. If you cite a URL and it 404s in a few years time, that's link rot.
So a DOI is a link that points to the item. It 'belongs' to the publisher who owns the article. They are obliged to update the link to point to the new location when it moves. Hence the DOI link always (in theory) works.
From a user's perspective, a DOI is just a URL that you visit that redirects you to a different URL. From inside, it's a link resolver system backed by a database that publishers keep updated through a registration agency (e.g. Crossref).
Different Registration Agencies do different things. Crossref has a tonne of metadata, contributed by publishers. Have a play here: http://api.crossref.org/works
There's a good Wikipedia article: https://en.wikipedia.org/wiki/Digital_object_identifier
But what happens if doi.org goes down? http://crosstech.crossref.org/2015/01/problems-with-dx-doi-o...
Most readers on the blog know all about DOIs so that section wasn't meant to describe everything about them. I welcome feedback though.
What happens if doi.org goes down? In the first instance you can use an alternative resolver, e.g. http://dx.doi.org/10.5555/12345678 can be resolved against http://dx.crossref.org/10.5555/12345678
In the second instance, we're as open as possible about it and discuss the shortcomings of the system! It's not perfect, but nothing is. We think it's the best solution.
It needs to be understood that "canonical" was really introduced in the context of search engine optimization and is supported by web search engines. In essence, if a landing page would have a "canonical" link to the DOI RL, the content of the landing page would be indexed under the DOI URL.
For browsers, the desired behavior would be to bookmark a page that has a "canonical" link under the canonical URL instead of the page's URL. At this time, to the best of my knowledge, this is not supported by browsers. However, it seems to me that it should not be impossible to achieve this, if, for example, the Persistent Identifier community (DOI, handle, PURL, ark, ...) would lobby for it. Alternatively, one could write an RFC and define another link relation type specifically for this relation type between landing page and persistent identifier. And lobby browser manufacturers to implement the "right" bookmarking behavior for it. But, the difference between "canonical" and this special-purpose relation type might be too subtle to make it through IANA registration scrutiny.
In the context of scholarly articles, if you're reading an article, the references will be linked via DOI, e.g. here: http://journals.plos.org/plosone/article?id=10.1371/journal....
Where this all gets interesting is non-traditional scholarship, e.g. sharing on social media where there are lots of 'publishers' (twitter, reddit) and they don't particularly care about best practice from the scholarly publishing industry. How do you get someone to share the persistent link? You can't force them. It's not a solved problem, and probably never will be.
Herbert Van de Sompel is working in this area http://public.lanl.gov/herbertv/home/ with things like canonical link headers that indicate the canonical URL of a given page. Here's a presentation he did on the subject, which he calls 'signposting the scholarly web': https://www.force11.org/video/signposting-scholarly-web
Is it just so they can try to extract some SEO magic juice from the URI?
As to why people don't... I wish I knew.