Yes, 100% agree. Several safe treatments that have (shown) promise should be approved conditionally, and approval revoked if trials/real world data don't pan out...
But no, let's take months to collect data so that we know what might have worked when the peak has long been over...