The aerodynamics of flapping aren't actually very well understood, so one of the reasons is just a platform for applied experiments. The reason you would want such a platform is because bird fight is potentially more energy efficient than anything humans have developed (as seems to be the case with so many naturally evolved mechanisms).

Noise? Drone propellers are ridiculously loud.

Ornithopters are more manoeuvrable than fixed wing drones, and (potentially) more efficient than quadcopters or helicopters.

I think the answer would be "what birds do is quiet and efficeent, we should do that!"

But really it's probably just like robotic hands - we just don't have small powerful/precise robotic muscles and we won't do a good job. So I think we're not going to completely emulate a bird soon.

Surveillance drones. If not in use in Ukraine already, they probably will be soon.

evolution produces end results that are worthwhile fully understanding at the very least, i imagine hovering around on updrafts is pretty energy efficient off the top of my head too