Per today's developer meeting, we decided on a 15 Dec 2011 release date for 2.3. The only remaining outstanding tickets concern documentation, and we want to ensure that the documentation undergoes some kind of quality control before the release. Unlike the code itself, which can (and will) be submitted to automated answer testing, the documentation will need to be tested by actual humans. We decided the best way to go about this would be to have a potential pull request acceptor first test the documentation by reading it and attempting to use the new feature before accepting. This shouldn't take too long, and should ideally be done using the acceptor's own data (to ensure that at least some modification of the script is taking place).
If there are any questions or comments about this, let me know. If people not in the meeting this morning are satisfied with this, we'll just adopt it as we go to close the tickets sam has added here: