Interesting. The video clips are especially valuable for understanding their methods. So, these tests have been done... but what other sorts of learning tests might be possible? Are there any that might even show value in this species beyond just "what is possible for the bee nervous system to accomplish?"
You know, there are repositories of publicly available image/video data from coral reef surveys.... somewhere online. They make them available for people to devise creative ways of pulling data out of such raw material. Might be something to look into.