That's a great link, thanks. Journal editors should hire technical editors themselves. This is easier than it sounds since they can just contract with a well-known company at competitive prices. The workflow: first, subject editors filter and revise submissions, and then scientists judge the merits.

There is great need for a software nonprofit to enable scientists and editors to collaborate with minimal overhead, because even $2000/paper is too painful. The true marginal cost should be under $200/paper for scientific editing,[1] and at that price point, open access will overwhelmingly win. If we can build such a platform, I believe that Elsevier will find it very difficult to retain their journal editors.

[1] http://www.biomedicaleditor.com/pricing.html