And, while we’re at it, another faulty argument that bothers me is the comparison of drunk driving accidents to high driving accidents. This is exactly like when the claim is made that “guns are safer than toasters” or some such thing.
What you need to compare is the conditional probability that one will get in an accident given that he’s high vs. given that he’s drunk. The fact that there are more accidents caused by alcohol isn’t even worth mentioning because drunk driving is probably a million times more common than high driving.