After selling 10-20 Million drones into the US market over the last decade the number of collisions with aircraft is countable on one hand and the number of fatalities remains 0. In contrast, between 300 and 400 pilots, passengers, and people on the ground are killed by manned aviation each year, and have had that rate for decades.
While there is a theoretical risk there seems to be no practical risk evident, particularly compared to weather and birds.
The main risk that drones pose is they are capable of reaching above the walls of sports stadiums to allow those without tickets to watch the games. In the US they are considered a risk to corporate livestock farmers who raise their animals in horrific conditions, release animal waste into local rivers, and otherwise commit terrible acts, but are located far enough from the roads that this can only be seen by drones. The claim there is that al Qaeda might get info on rural hog farms and use that information in some way. No one says what way, but it's presented as a terrorist danger.
There was recently the case of an engine suffering what appeared to be compressor stall; at least it was belching flames in a spectacular manner. This was attributed to bubble wrap. Which is one more case of bubble wrap doing damage to an aircraft engine than the 10-20 Million drones have managed in the USA, which is 0.
Can stuff happen? Sure. A seat switch could put a dozen people in the hospital. Which is also a dozen more people than drones have injured with respect to aviation operations in the USA.
I'd rather make active avoidance a key component over depending on some long chain of rules that are certain not to be followed by all users.