A new publishing paradigm

Anyone that has been working with science has thought about the publishing process and the pitfalls that goes with it. So, I will not list those here. Instead I would like to propose an alternative to traditional publishing.
The main idea is very simple: instead of publishing individual articles, we should copy the modern software development process more or less straight-off. Someone, or several, maintains a "publication". This is simply a git (or some other vcs) repository containing something resembling a book. Any scientist that has something to contribute to the book can send in a traditional pull request to complement the publication.
There will be so many advantages with this: less writing for each published unit, more consistency, an updated modern and efficient peer review process, open access, simple referencing (by making versioned releases or referencing a commit hash).

The next step for the arXiv

The axXiv has become a standard place to put up pre-prints of journal papers. This is great with increased availability and distribution speed. But, it could be taken one further step. Add an automated peer-review process to the papers, and the system is ready to completely replace the blood-sucking bandits of publishing houses. This is money the science community can use better, and we are anyway doing all the work: the research, the paper writing, the peer-review. We could be adults and take full responsibility for the whole process and improve quality the same time.

So, my proposal: develop a review AI that does the first screening, checking things like language, style consistency and formalities such as affiliation, contact information and plagiarism. Next, the system can propose reviewers together with the authors. The system can also propose an editor that can do the work of the editor on a per-paper bases.

Wikimedia foundation the ideal publisher of free scientific papers

The issue of open science is a pressing one. The large publication houses, Elsiver, NPG and the others make huge profits publishing scientific papers. And the situation is actually absurd! The publishers really do nothing - they do not write the papers, perform the research, do the review, or the typesetting (ok, maybe thy do typesetting, but really it is to a minimal degree with the modern typesetting systems).
There is a slow drive for so called open access. However, this is something that even further increase publishers income since open access papers are funded by huge fees from the publishers to the authors.
So, we need some new actor stepping in and taking publishing responsibility, including review assignment, This is really the key problem that needs to be solved in true free science. How to manage the review process.
My suggestion is to have a process similar to how Wikipedia works. Maybe in combination with other elements such as a score system similar to Stack Exchange. And really Wikimedia is the ideal organisation to take on this task. It is the natural extension of their current operation, maybe in connection to Wikiversity. But a completely new project would probably be even better.