It is weak evidence for the effectiveness of hydroxychloroquine. e.g. if you ran this as a Bayesian analysis you would find that these results should shift your beliefs towards it being more likely that hydroxychloroquine reduces deaths + ICU visits.

However you are correct that the medical community has decided to be conservative to reduce the likelihood people make decisions based on weak evidence. That may be the correct decision from a overall social welfare perspective. In that framework all we can say is "this experiment did not generate enough evidence to reject the null hypothesis".