To get reliable statistics I suspect you'd have to repeat the experiment literally thousands of times across a representative sample of ADB hardware (not just one random Mac model that may be more or less susceptible than average) and also explore a range of auxiliary variables as well. (For instance, does the angle of the plug as it's inserted, thereby changing which wires make contact first, make a difference?)

The one thing we do know WITHOUT experimentation is if you don't hot-swap you won't blow out your port from hot-swapping.