Discrete events are a human construct but the underlying data is continuous. I'm not sure the attribution of individual events is that meaningful and the articl

Discrete events are a human construct but the underlying data is continuous. I'm not sure the attribution of individual events is that meaningful and the article suggests for a variety of reasons the models aren't really built for that purpose anyway. So I think in a sense you are right, but this doesn't necessarily imply any problem for the predictive value of the models or research going into them.