One problem is that people sometimes learn the wrong lessons, or draw the wrong conclusions after a storm. Normalcy bias also leads people to discount the possibility of a bad storm, but patterns of not so bad ones make them expect that and not consider what is possible.

Burncycle, while Florence became weaker as it crossed the coast in terms of wind speed, the slow forward progress and copious rainfall brought the bad flooding that a faster storm would not have. I think they need a new paradigm for classifying hurricanes that includes all possible effects, not just wind speed.