I'm not sure what kind of strawman you're building here. Are you thinking that I'm arguing that all or even most norm-breakers are proven correct?
I am arguing only that some norm breakers turn out to be prescient and thus overconfidence based on norms alone is treacherous and bound to be proven wrong at times as we learn more, and Normalcy Bias often prevents people from adequately assessing that risk. Here's another wikipedia link!