I stand corrected, thank you. Furthermore, the journals seem to have many more articles than I thought - I had estimated 20 per journal, but most PLOS journals seem to have 30-40 articles. In my original $ estimate I also assumed 10 issues a year, most PLOS journals seem to do 12.

So consider $2250/article as a representative price in the PLOS suite, by 35 articles by 12 issues - About $USD950,000/annum revenue per journal. I wonder how that compares to revenue/issue for the traditional subscription model augmented by online fees? I looked at Elsivier and Wiley's financial statements but there wasn't enough info broken out to guess.

EDIT: closer reading of Wiley's statements [1] suggest $1.385 billion revenue from journal subscriptions with 1600 "scientific and professional journals" in print, for revenue of $865,000 per journal.

[1] http://www.wiley.com/legacy/about/corpnews/fy15_10kFINAL.pdf