the stale pages and spy have been resolved. there are some issues with queues not being emptied quickly enough that i think are database transaction issues -- we're submitting more database queries and modification than previously by an order of magnitude, but the database is still running on a single process
the layout issues are complex, and weren't due to a code change, which is extremely odd. we're also currently without a dev environment, so it's difficult to make large changes as they can't be tested before deployment. i'm fixing those things as i have time, but in the meantime, weirdness will happen. that said, i have repro of the layout errors, so i shouldn't need more reports to debug. again, it's the lack of a development environment that's holding rknize and i back at the moment.