Personally I'm very much for safety devices that test and monitor those operating a machine to check if they're doing so safely. I'd also test humans who want to operate various machines, particularly cars and other motorised vehicles, for a whole set of, let's call them, inclinations - those that tend to cause "accidents". Prevention is better than cure.
Some safety devices are not preventative of dangerous behaviours but rather preventative of the consequences of such behaviours. Seat belts and air bags in cars are like that, as is the Sawstop mechanism. They don't prevent dangerous behaviours by a human but rather reduce the consequences of such behaviour. In some, they actually increase dangerous behaviours. That risk compensation effect.
Its a subtle difference between safety devices that prevent risk taking rather than reducing risk consequences.
Those devices that reduce risk consequences are also those that tend to induce the risk compensation effect. "I'm now safe so I can take more risks". This would be fine if the risks taken produce only reduced consequences for the risk taker. With cars and similar the risk consequences actually rise for those outside the car, as a risk-hungry driver ofttimes turns his vehicle into a projectile.
Sawstop is a consequence-reducer rather than a risk-reducer but one that doesn't pass any consequences of greater risk-taking by the user to others. In that sense it can be seen as a very positive safety device. The additional risk, though, seems to be one of inconvenience and cost when the protection mechanism operates. A small price to pay compared to a finger or hand loss, especially in the USA where one may also be driven to penury by such an event. But still an ongoing cost, especially if it fires with false positives.
So why not consider if there are other ways possible that will reduce the ongoing costs and the rather large inconvenience of having to rebuild the safety mechanism, replace the blade it damaged and hope the large braking forces involved haven't damaged the arbour or the motor? Are such methods impossible; or is it more a matter of a jealous patent owner shutting down any and all alternatives?