The set screws and gears everyone is blaming don't play any part in the operation of the sensor when the A/C is in flight. They keep the sensor in a faired position on the ground.
Unless the screw falls out and jams the gears?
There is an interesting design point here; nothing in the real world is independent. You would think that the ECU of an electronically controlled engine would be completely independent from a marine hot water heater, or that a freshwater toilet and an autopilot would have nothing in common. Yet, a small leak in the cooling system circuit (which is used to heat the hot water under way) dripped down a wire loom and into the wiring of the ECU, causing unexpected acceleration at random times depending upon which set of throttles was being used to control the engine. The rate compass which was used to set the ships' course for the autopilot was installed just a little too close to the DC pump for the freshwater toilet, leading to a magnetic field interference issue that caused large swings in course when the toilet was flushed. There are examples everywhere; at my sister's workplace they were prohibited from changing the screen color on their work PCs because it would cause their application to crash (back in pre-win95 days.) One NT bug that I never figured out at work had to do with setting the date to before the 1700s which caused death deep in the operating system only if a certain application was in use at the same time.
However, the managers still love to draw little boxes on the whiteboards to show how simple it all is and how any idiot can react within five seconds...