The time of administration matters as well. A lot of the early studies were made in high-severity, hospitalized cases. At that point, the results were mediocre

The time of administration matters as well. A lot of the early studies were made in high-severity, hospitalized cases. At that point, the results were mediocre if not totally ineffective (lopinavir/ritonavir, for example). It might be better early on, but there is yet a lack of trials for early administration of drugs. Stuff like the RECOVERY trial in the UK will be pivotal in telling which treatments work and which don't. Also I don't mean only repurposed drugs (aside the very debated HCQ, there's ivermectin that is worth at least looking at), but also new molecules made specifically to fight the disease (there are quite a few slated to start clinical testing this summer). Then "underdogs", like favipiravir, which apparently cuts the hospital stay time in half, but the few studies published aren't that great (no placebo, control is standard of care which means everything and nothing). I would also add that the lack of standardized protocols hurts a bit the interpretation of the studies. Take heparin for example, which is believed to drastically cut mortality (although the lone study published is full of confounding factors): different hospitals in different countries have different protocols of administration, further complicating things and making difficult to understand proper dosage, etc.