How would anyone know? Which is the point. Many drugs save one group of people and kill another. Like penicillin.
I wouldn't want to be responsible for making that decision for someone else. Many patients who are severely ill and having difficulty breathing are surviving without that drug.
If a large group of people in one city could be (all) given the drug under the same clinical conditions (that takes some definition), then in another city of same size, no one got the drug (I'm sure we could find one), that would help sort it out.
In those places where people survived (after the drug), I'd like to see the mortality rates from the same hospital for people who also got the drug. If it is the case that no one died after receiving that drug, that would be quite a recommendation even given its lack of trials. But if that were the case, then the doctors involved would (I believe) be pre-publishing the results all over the place. Writing letters to journals. Having pressers.
I find it hard to believe that all people who have been giving the anti-malaria drug have lived and the only way we know about that is basically through SM. Instead, there are some people who died and it is their conditions, allergies, ages, etc. that we need to know before we give the drug to everyone.