You're playing the role of skeptic, but you're not going far enough. How would I actually answer that question? You could have said the same thing about the Therac; of course it can't blast that much radiation, there's no good reason for it. Yet, here we are.
What evidence do you have that it is actually not possible for the machines to dangerously malfunction, and how much confidence can actually be placed in it? Those telling us it is safe don't have entirely clean hands on the topic, they've lied about other things like picture retention policies and capabilities before.
You're making an appeal to incredulity. I would not actually be that incredulous to discover that the machines do indeed have some sort of flaw that would permit them to blast out more radiation than intended. After all, it's happened before. That said, I don't necessarily consider it likely, either, what I find far more problematic is that this is merely one consistent piece in a larger puzzle of unconcern about the real safety of the public.
For the same reasons we ought to be able to examine our voting machines, we ought to be able to examine these machines.