If you think about it long enough, that is irrelevant. It's just a number that is close to human reactions. Move it to the actual study, and once someone gets DQd by 0.001 seconds, then the standard deviation in the study will need to be part of it. And the closer you get to the natural reaction time, the more likely you will get runners trying to guess the time.
The start is something they practice a HUGE amount, and for professionals, making it 0.1000 seconds is pretty clearly possible and very likely. A gross majority of competitors don't false start, afterall.
If the runners universally had a problem with 0.100 seconds, let them take care of it. Otherwise, they fully accept the time and the consequences.