There seems to be a great deal of misunderstanding among Wikipedians how academe actually works. Piotr thinks a grad student can produce a scholarly journal. Look at history. In reality it takes hundreds of scholars working together (almost all of whom are paid professional salaries by universities.) Printing and mailing costs are only a fraction of the total expenses for a scholarly journal, so the advantage of going electronic is small in terms of production costs.
I talked just now with the editor of ''The Journal of American History'' --I used to be on its editorial board. It has dozens of editorial board members and hundreds of unpaid scholars who evaluate articles and write for it. They are paid not by the Journal but by their own universities to do this kind of high prestige "service." (History professors are paid for research, teaching and service--the average salary in USA for a full professor of history is $83,000 plus 25% benefits.) The Journal has 14 in-house staff members, who are paid salaries at rates standard for Indiana University. Most have PhD's or are PhD candidates--that's eight years of specialized, expensive post-graduate education. Book reviews are a main role. They read 3000 new books a year and select the most important 600 for actual review, using a database of 11,000 available scholars. 300 full-length manuscripts a year are submitted and the senior editors and outside reviewers narrow that to the best 10%. The staffers do intensive quality control on the accepted articles and are backed by a major university library (which is expensive.) They occupy nice offices with phones & computers etc that are also paid for. The Journal pays travel expenses for meetings. The output is 4 issues a year with 1300 pages of high quality scholarship delivered to about 10,000 historians and libraries.
Indeed anyone can try to publish a junk history journal single-handed and give it away free; almost nobody does so. The software is there but the necessary expertise is very expensive and takes decades to develop. It costs real money to produce the "reliable secondary source" that Wikipedia wholly depends upon. The question is who pays for it.
Richard Jensen