We seem to know enough from the article, however: (i) to test the "gateway" hypothesis, a longitudinal study is required, since time is needed for the drug user to get through the "gateway;" (ii) the study has focused exclusively on teenagers, so it is safe to assume it was not longitudinal.
To say anything on the subject, either way, is therefore over-reaching. And, what is even more disturbing, you see a pattern of such over-reaching. Observe that from a study of one drug use, you cannot draw any policy conclusions. Drug use is sociology, chemistry, and the like. But policy is a managerial issue: one needs to know the costs, the priorities, available alternatives, etc. Thus, the author gets into the area beyond his qualification.
To see it more clearly, consider an analogy. Suppose an engineer makes a study of accuracy of a particular weapon, and then gets up and tells us that we should reduce the use of this weapon by 30%. You would laugh, of course: accuracy of one weapon is one issue, but its role in the arsenal of the country depends on what alternative weapons are available; on the costs of production and not just the accuracy; etc. The use of the weapon depends also on the military doctrine, the strategies uses by teh generals --- something that the engineer knows nothing about and certainly have not addressed in his study.
Well, that's why you never hear such nonsense from engineers. But in social "sciences," it is somehow acceptable to do so: make a study of one ant and then make recommendations for the rest of the universe. That's what this author is doing.