Yes, this is part of the equation, but you also need to balance your efforts based on the "risk landscape". Spending effort in one place generally means that less effort is spent elsewhere. I think it's a direct parallel to optimizing code: measure first, then optimize. Making your inner loop 10% faster will make a difference, while speeding up your logging won't. Or vice versa, if you have no single inner loop and your logging is abysmally slow: measure first!

If it turns out that drones have 1000x the per impact likelihood of causing a fatality than impact with a birds, but that you are 100000x more likely to hit a bird than a drone, then are you better off spending your efforts on reducing the risk of drone impacts or bird impacts? Or (seemingly perversely) is it better just to tolerate both risks as they are if measurements suggest greater return by concentrating your efforts on improving airport security (or mental health checks for pilots, or testing English proficiency, or even traffic safety on the way to the airport).

In this case, my guess would be that adding regulations to reduce the risk of accidental drone strikes by aircraft will have negligible returns, that common sense "don't fly near airport runways" is sufficient, and that there is lower hanging fruit that makes more sense to concentrate on. Normal Accidents by Charles Perrow is an excellent book if you are interested in this topic: https://en.wikipedia.org/wiki/Normal_Accidents