You have the math backwards. It would be like saying that someone who guesses your 4 letter pin code is a mind reader because the likelihood of that is low.
You are not factoring in how many times he has tried to guess other people's pins. Or if some pins are more common than others.
Similarly here you are not factoring in how often an entire group of sensors has the same big and repots incorrect data at the same time in the same way. You are Cherry picking the one time the data looked like it might be a UFO. This is also a form of confirmation bias.