Assume for a moment your assessment of remdesivir is spot on. It does nothing. Still, approximately 80% of patients hospitalized with COVID-19, and 60% of those admitted to intensive care units (ICUs), survive. So with an ineffective treatment a majority survive. (1)
Now apply MO's model, accounting for the fact that most ICU patients survive.
In a small sample set, 1 clinical test site, it will be very difficult to determine with certainty that an experimental treatment is having a positive outcome. Particularly in a blinded protocol.