(thanks for adding that here, Tree).
I now think the culprit is the new content caching system that we introduced in 1.5.2. I couldn't spot the problem in the code or reproduce it locally, so for now I'm just going to comment it out, which means things will be a tiny bit slower (imperceptibly so except on especially involved pages) but this problem shouldn't arise any more.
I have made this change and deployed to our testing branch and will try to deploy it live by the end of the day. If you see any more problems, please report them here (or on another support ticket -- whichever is easiest).
At the same time, I made another change that will allow you to clear the cache yourself when you see oddities like this (if you have the admin role). See add url to clear cache. I should note that John Abbe has been prodding me to do this for a long time, and I've resisted thinking it might be challenging. It took about 5 minutes to build. Sigh. That change will be deployed at the same time.
OK, it's deployed. I'm going to be optimistic and mark this ticket "answered", but obviously let me know if you see anything else.
--Ethan McCutchen.....Tue Feb 15 15:34:34 -0800 2011
Tree, I added a link to "Clear the cache" to the Administrator Links card in our sidebar. So if you see similar problems in the future, try clicking that and see if that clears things up.
--John Abbe.....Tue Feb 15 16:18:10 -0800 2011