The majority of cannabis legalization has occurred in the last few years. The paper you linked uses decades old data and provides statistics on overall inciden

The majority of cannabis legalization has occurred in the last few years. The paper you linked uses decades old data and provides statistics on overall incidence of schizophrenia with varying durations of prevalence. Bayes rule applies here - not everyone will go buy cannabis and use it daily. Unless a study is accounting for these factors, it's useless in making a determination on the connection between psychosis and cannabis. Your own point applies, frequency must be adjusted for, by an understanding of priors.