When switching editors, we re-fire the ready and loaded events, with timing since the very first load (so potentially minutes). This data is not very helpful. If we want to measure the performance of switching we should use different event names.

edit: Looks like we just re-fire the loaded event, and ready is already correctly ignored.