Fundamentally what is needed is a process that produces peer-reviewed, immutable, discrete pieces of globally published work. Self-published blogs don't really achieve that well.
Institutional archives do.
I don't know that you're wrong, but:
Right now, I really want an article in Nature or some other prestigious journal, in order to advance my career as a scientist. If those prestigious publications were suddenly gone, there might be other publications that would spring up to fill the vaccuum, but the new ones would have no track record and no prestige. So wouldn't everyone just as soon either self-publish or else favor non-exploitative, open-access journals?
What if we did away with all car companies? How about all TV media companies? What if all physicians were no longer physicians tomorrow?
From that perspective, I am assuming that the top (in hierarchy, not importance) groups from the then-incumbents will be the fastest to raise new companies, which will still be able to attract talent with the prestige that the old company used to have (think about how startups founded by ex-FAANG work). E.g. Nature executives and leaders develop N2, FOX builds Wolf, Tesla builds...Tesla, all of which start off with more credibility and popularity than offerings without this benefit. I do not think that the exact same situation would arise though. There would still be easier entry into these markets than before, at least at first.
In other words: for many disciplines, the potential replacements providing peer review, archival and distribution are already there. The main thing a replacement system will have to find out is how to determine who to give grants, tenure, etc.
(Disclaimer: I'm working on a project that hopes to achieve that.)
It's not currently that interesting to the average HN reader though. The idea is to allow researchers to give each other recognition explicitly, ideally replacing the recognition of "being published in journal x". Since researchers already have a reputation in their academic community, we figured that we could leverage that as a counterweight to the indirect reputation that journals currently have.
We do so by integrating with preprint servers and publishers, providing a widget that allows researchers to endorse the research they're reading. The first integrations are set to launch at the start of 2019, so until then there's not much to play with.
That said, the PageRank-like system you might be looking for could just be the h-index? https://en.wikipedia.org/wiki/H-index
However, for academics, it wouldn't be entirely clear what to do. For example, if you choose to only disband paywalled journals, they will likely simple submit their works to open access journals, and over time some of those will emerge as the new "good ones".
Which is an interesting parallel, because that is sort of what Plan S tries to do. Except that it doesn't really have the power to "break up" paywalled journals, but only to deny them the ability to publish work they fund. As more and more funders join (like these Chinese funders appear likely to do), the closer this will get to "breaking up".
If you were able to disband all journals, but not preprint servers, then researchers would probably start publishing their work there, new servers would emerge for disciplines that don't have one yet, and some new system of determining "excellent" work will probably emerge. (Disclaimer: I'm working on one such system.)
If also preprint servers were disbanded... I don't know - nobody would be able to publish any more? Would you allow new publishers to be created?
Yes, that was my intent. New publishers could emerge, and with no new restrictions on their behavior. Basically the main change is that the slate will have been wiped clean in terms of prestige.
I'm wondering
1. would new publishers emerge?
2. how, if at all, would they be different from what we have now.
These scientific journals hail from a time before the web was widespread and universally accessible.
In the present day, it is technologically feasible for every scientist to self-publish on their own blog (we'd probably want some new tooling for tracking and counting citations distributed across the web, but that's just details for some programmers to work out).
[QUESTION:]
Given the premise, I wonder the following. If we were to simply disband the existing publishers today, rather than trying to regulate them, what system would emerge to fill their place? Would it be similar to the existing system? Or would it be something more free-and-webby as described in the premise? Or something else entirely?