For small IOT devices, the primary benefit would be to charge from a distance (wireless charging) but then the efficiency of the ultrasonic energy transfer will also drop inverse square to the distance. And if the power requirements are low enough, why not live with a lithium batter that will last 5-10 years possibly?
I can think of some specialty environments where batteries fail (due to environmental, temperature, acceleration, etc.) but where energy-transfer would be good. I can separately think of times when RF/induction/etc. wouldn't be good, but ultrasonic might be good.
But I guess it might be useful for other things? I dunno.