None of that is true. You can have a d10000 critical chart full of interesting results. Why are you suggesting that "interesting" is the same as narrative mitigation?This just seems to contradict your previous points. If outcomes should be random, why are you narratively mitigating the result of a "fell where it may" open die roll? If you're going to salvage the fiction with "interesting outcomes," why bother rolling in the open? All you're doing is teaching your players the dice don't matter when it counts.
This is less an issue in storygames that focus on "failing forward," but I thought that was the very phenomenon you started the thread to argue against. Random outcomes mean sometimes the worst befalls.