Someone paid by the publisher needs to look at every paper.

Some journals calculate internal costs of up to $40,000 per paper (see http://www.nature.com/news/open-access-the-true-cost-of-scie...). You seem to be skeptical of such claims, but simply asserting that the expenses don't exist isn't enough.

> Someone paid by the publisher needs to look at every paper.

Is that true? If so, what is the use of that? I don't see any scientific added value. I never even noticed such an activity myself. We can get rid of it where it still exists.

> Some journals calculate internal costs of up to $40,000 per paper

$40k per paper ? Haha sorry but I just don't buy that.

I'm not asserting these expenses don't exist. However (see my other comments in this thread), I'm showing that you can avoid them, and how the same quality control of publication can be obtained without having publishers around. And in the example I took close to everything is done in the traditional way. Which in turns is a demonstration that publishers added-value is minimal.

>If so, what is the use of that?

To prevent bad papers from getting through. Editors at a paper are the first filter, peer reviewers are the second.

Again you are not right about that. In practice in the vast majority of the cases¹ the editorial board (for journals) or the program committee (for conferences) do that. And these people are researchers, and they're not paid by the publishers. At this stage the publisher does not intervene.

¹ In some other cases such as with Nature, what you said is true. But the point is that the model you are describing is far from being the only one, and there is no evidences that it is better. In computer science for example, the most prestigious venues are conferences not journals, and in that cases the publishers does nothing until the very end of the process, where its work often consists in putting PDFs online and that's it.

Computer science and math are very different than other fields, because they can stand independently most of the time.

If you're going to throw around phrases like "vast majority", can you cite numbers?

I do not have precise numbers, but I've worked on these issues with biologists, philosophers, physicists, computer scientists, mathematicians, and people in cognitive science, and also with librarians and people in charge of negotiating with publishers. What I told matches with what all these people also see from their point of view.