You don't counter "heavy survivorship bias" as you frame it by swinging to the other end and talking to ER docs about their horror stories; humans don't process

You don't counter "heavy survivorship bias" as you frame it by swinging to the other end and talking to ER docs about their horror stories; humans don't process information by consuming extreme data points and then internally averaging them into some sort of model. You look at the data in aggregate and end up pretty much where the experienced touring posts here land, which is rare but real risks sprinkled across typically accident-free riding. "statistics is against you on this" is NOT the reality, in terms of likelyhood of accident or severity comparing car travel to bike travel.