Klaus,
It's an interesting subject. It would seem that the insect has also signals coming from the different character of air blow back, (from wing movement), from the end of the petals versus the inside of the petals and might be able to locate itself in 3D by sensors on its hairs, (like the hairs in the mammalian eustachian tube for hearing). Of course, I have no idea whether or not insect hairs have sensitivity, but it seems feasible!
I guess one can freely speculate and then that gives questions that can be checked against know facts or allow new experiments.
Asher