I'm not sure we have any disagreement, Doctor580. Small sample requires large effect. Small effect requires large sample. What I'll add to that is that because most m/m patients recover fine within 14 days, there's not much space left to put a large effect size into.

With the severe/critical, we'd need a striking effect to show statistical significance with n = 50 patients. For example, remdesivir reduced the mortality rate by about 40% (1 - 0.071/0.119) compared to placebo. Sample size was over 1000 but still no statistical significance.