Not disagreeing. I am a physician who runs clinical trials so I have perspective.

Agree that trials are needed, the VA study was not a legit, prospective controlled trial. That is the point and the problem. By the way, the notion that mild patients don't need treatment is not quite accurate. A drug that reduces hospitalization from 10 to 5 percent makes a big difference in morbidity and mortality and also likely predisposes us to faster herd immunity which breaks the chain of transmission.

Also the fact that presentation is so varied likely needs a huge sample size (thousands) to show efficacy unless the drug is dramatically effective in most patients. Leronlimab may fall in this category.
Otherwise we need to properly match controls which may not be possible here. Age, sex, ethnicity, and medical history may not cut it. Unlike an HIV transfusion, the majority exposed may not even test positive--there are factors we don't know about, immune system, lung cell receptors that might be at play. Without knowing that information, I am not even sure we can run a proper controlled study Even a positive drug result may be a false positive regardless of P value.