I suppose the problem with regulations is how do you word them to take into account the improbable but obvious. NASA refers to "criticality one" to describe components that, if they fail, the rocket, shuttle etc. will crash.
Anything that flies has, and will always have, lots of items that can be concidered "criticality one." For airplanes, obvously the wings and tails are at the top of the list. For helicopters, blades, masts, gearboxes and even tailbooms are a few such items.
So how does a regulator write language to take into account items like those that are "criticality one?" I would suggest that regulators, like engineers and managers work in a world of compromises. Just as an engineer must design a system that is best possible, by definition he will make compromises, the regulator will write language that will try and make the system the best possible but, in the end will also make compromises.