Pawel Szczesny at Freelancing Science has a post considering models of open science, and how best to encourage it. For one thing, he says, it would be helpful to test various models outside the scientific arena: "Principles of Open Source software did not prove to be useful in open drug development. ... Crowdsourcing will not advance quantum physics," he writes. "Open Science in its fullest form is not an issue that scientists can truly solve by themselves."
In a post he advises readers to "take with a pinch of salt," Neil Saunders has a post on social bookmarking options offered on journal websites. PLoS and BioMed Central clearly get it, he says, but a snapshot of the bookmarks available at Nature (limited to Connotea) leaves Saunders asking, "So NPG: is it 'we promote sharing' or 'we promote sharing so long as you use our service?'"