An odd spinoff of the ongoing debate over at http://rjlipton.wordpress.com is a proposal that the genius logician Harvey Friedman (first taught as a prof at Stanford in philosophy at age 18, a world record for youthful professoring) has put forward in the middle of it for refereeing papers at journals, obviously inspired by the ongoing, now many hundreds of entries long, debate over the proposed proof by Vinay Deolalikar that P does not equal NP in computational complexity theory (current consensus: current proof flawed, but argument might still be right, or more likely, proof strategy may be very productive for lesser results).
So, the proposal is that an author is offered the option of public refereeing rather than the standard secretive double-blind type usually done. This public refereeing involves the journal putting the paper up on a website where anybody can publicly critique it. The author can respond and put up new revised versions (and can criticize the editorial board of the journal as well). When the author is satisfied with what has transpired out of the process, s/he can propose that the ed board consider it for publication. They then decide either to publish or not to publish. If it is not published, the last version can either remain up "hanging" on the website or be taken down.