I have to say, I am really a fan of the "Letters from the Editors" blog that Jonathan Jacobs et. al are running. I think it not only provides a great (and long overdue) inside look at the academic publishing process, but that they are also opening up some important issues for discussion.
For instance, in the most recent post, Carrie Jenkins (an editor at Thought) discusses an interesting question by Wesley Buckwalter on conflicts of interest. Basically, Buckwalter notes that although reviewers are expected to report any conflict of interest they might have in reviewing a paper, there is no disciplinary-wide consensus on what comprises such a conflict, or when such a conflict might warrant declining to review a paper.
In her reply, Jenkins focuses on conflicts that might compromise anonymity (e.g. knowing who the author of the paper is). Then, however, in the comments section, Jonathan Ichikawa Jenkins notes that not all conflicts of interest need compromise anonymous review. Ichikawa Jenkins notes, in particular, that one may face conflicts of interest in (1) agreeing to review work that is deeply critical of one's own work, or (2) agreeing to review work similar to unpublished work one is working on oneself (I commend him, by the way, for noting that he normally declares these things to editors!).
Anyway, although it is of course very difficult for editors to effectively enforce norms for disclosing conflicts of interest, I think these are very important issues to discuss and arrive at disciplinary norms for. My own sense, for what it is worth -- and it is admittedly anecdotal -- is that some people play far too fast and loose with these things. One time, for instance, I actually had a reviewer forget to delete his name from the "properties" section of the review document he submitted...and, you guessed it, the grounds he gave for rejecting my paper were none other than that the paper contradicted his own position on the subject (he even went so far as to call my project a "nonstarter", without any explicit justification). The way I see it, there is no way he should have agreed to review my paper if that was the kind of review he was apt to give it -- but again, there seem to be no well-known, settled standards for these kinds of things.
Anyway, what does everyone think? What should norms for disclosing potential conflicts of interest be? (I look forward to hearing what everyone has to say!)