Many articles on Wikipedia are already conforming to your description if I understand it right. All scientific work can maybe put in something similar to the present concept of a “series”. However, what will the magnitude of that be? “All the scientific literature” is a very big entity. The exact complexity level for the “average audience” is also a problem. An encyclopedia is supposed to be comprehensible by any literate person, including older children. Some Wikipedia articles are scientific papers, full stop. I am also an independent “researcher” interested in evolution and specifically evolutionary psychology. I have no training in psychology however. My library do not have the materials I am interested in. Therefore I make use of Wikipedia and open access articles and have a reasonable idea of what Wikipedia presents and the quality of that vs peer reviewed OA literature. While some articles, as said above, are complicated full-fledged scientific articles, other stuff supposed to be scientific, is real rubbish – technically and content wise. I have seen better similar info on lay / popular websites. I have a good idea of the quality requirements of a scientific paper. I have experience in helping my library patrons with references when they review a paper and also struggled (or still is busy) with my own first peer reviewed paper (believe it or not, in yet another field I have little formal training in, history). I will still sweat blood before it is ready.

So, where Wikipedia is a great source of information for the advanced lay enquirer, I have learned to verify everything with what I can get hold of (occasionally also buying a PDF from the internet) before believing it. Your proposal may be good but will it not be possible to build in a peer review system in such a Wikipedia division, just keeping into mind that the language and style should be comprehensible. However, such an entry cannot go online with remarks like “citation needed” , it must be verified and corrected. How this will work out in practice with Wikipedia being a crowdsourcing system and with not much money, I do not know.

Wikipedia is an "encyclopedia" which contains articles on different subjects. It's not built to gather lists of publication. Instead of creating a list of literature, why don't we cite these publications in their related article? Take note that in every Wikipedia article, there will have section(s) like References, Further Readings, or Notes. These are the publications or resources which have been cited in the article. By inserting more information in the article as comprehensive as possible, with references cited properly, we can make sure reader to gain knowledge by reading through the article. Furthermore, we should actually educate researchers or students to always verify the content of Wikipedia by checking through its list of citations. This is the correct way of utilizing Wikipedia in research or other learning purposes.

Interesting question. I foresee great difficulties within such a project. 1) My experience in reviewing the literature on the problem of mechanisms of a folding formation has shown that a meaningful text arises when the researcher already has an idea of the evolution of views on the basis of his own experience; when the solution of the problem exists almost. It means that a good review (in order to provide information to fresh graduates) will appear an individual, compiled by an experienced researcher, rather than a collective spontaneous, as technologically provided in Wikipedia. 2) Any problem has its own "architecture", that is, the hierarchy of large and small parts of it: "postulates" - "main theorems" - particular problems, etc. This structure of the problem will be not necessarily generally recognized, it is likely to be the subject of discussion, and it contradicts the meaning of Wikipedia. 3) There is a problem of methodical literacy of much researchers; it means that not all the researchers can really distinguish between important work and the popular (most cited) one. Therefore, not always the most cited works are the most important in terms of the evolution of views on the problem. My experience in the tectonics leads me to the growth of skepticism about such methodological competence. In this sense, a collective effort to select works for review in such a project is likely to fail to provide the expected automatic "solution" to the problem. GENERAL SUMMARY – a good review of any problem can be made by a specialist in this problem, not by a team.