A real drug might really treat some subset of the cases of an issue, if some issue has multiple unknown causes and the drug addresses one of them. If a drug cu

A real drug might really treat some subset of the cases of an issue, if some issue has multiple unknown causes and the drug addresses one of them. If a drug cures 10% of people with an issue, and does not cure 90% of them, is it still worth trying out on people? Is it still worth researching to try to whittle down diagnoses to hone in on those 10% instead of just claiming it to be a failed drug?