After all, we do expect to be peer reviewed when we submit a manuscript—for work that we have already done.
http://plosbiology.org/article/info:doi/10.1371/journal.pbio.1000197#article1.body1.sec3.boxed-text1.sec1.p8

I am in complete agreement with the arguments and solutions proposed in this paper. I would simply like to underline that, similarly to people, it would also be much better if published papers were rated retrospectively for their own quality and impact rather than via the impact factor of the journal that their authors managed to "sell" their work to.