Oh, and I forgot to mention an important feature, which I often take for granted as obvious but is worth spelling out. much of the work that goes into processin

Oh, and I forgot to mention an important feature, which I often take for granted as obvious but is worth spelling out. much of the work that goes into processing a manuscript goes into processing all the manuscripts that come in, not only the ones that get published. So when you have a journal that is intending to serve as a filter for importance and is as selective as possible (ie Nature, Cell, etc), you end up rejecting 90+% of articles. All those rejected articles take work to process and come to a rejection decision. So the name of the game then is trying to cover the cost of all the things you reject by monetizing all the things you accept. That's why megajournals that intentionally do not filter by importance (and thus can accept a higher proportion of submissions) are able to charge less (it's obviously not pure cost economics, top tier journals charge for their brand name as well of course).