> “Overall, regular use of cannabis [...] was found to have harmful consequences, regardless of the age people began using it.”
I am skeptical that this study is actually representative of the effects of cannabis, and does not instead depend on other, highly correlated effects, which are then presented as the reason. From my general (non-user) perspective, I would assume that there is probably a bias towards groups of people who have a higher chance to use substances in general, or live in a comparatively poor environment.
For example: Children of smokers often become smokers themselves (as numerous credible studies show). If they now start with cannabis, they would be included as the victims of cannabis in the discussed study. However, the causality is probably reversed here.