Regardless of whether they're just fear mongering or not, from a statistical standpoint, it makes senses. They're just marginalizing over the uncertainty of the expert.
p(blackhole = 1 | expert opinion) = p(blackhole = 1 | expert opinion, expert=correct) + p(blackhole=1 | expert opinion, expert = wrong)
It's the type of calculation that happens every day. E.g., what's the probability that I'll die in a car accident given I have an airbag in my car
p(survive = 0 | airbag in car) = p(survive = 0 | airbag in car, airbag works = true) + p(survive = 0 | airbag in car, airbag works = false)
If you don't take into account the uncertainty of the expert (or airbag malfunction, etc.), the you're ignoring what could be a large contribution to the actually outcome.
Or you walk into the contract negotiations with a red pen and cross out the 12 month extension beyond employment termination.
The problem with non-competes is that they are so vaguely written that it can cover pretty much any new job in the field you're trained for. Sure, you can bring it to litigation to prove otherwise, but what employer would be willing to risk taking a new employee that might have a possible contractual obligation?
"Love your country but never trust its government." -- from a hand-painted road sign in central Pennsylvania