As a layman, my main observation with the link is they seem to be computing a statistical measures over the studies to get a meta-result. But that isn't necessarily a statistically valid methodology, for the following reason: it is like if someone gave you a bunch of grapes and and apple, and asked you to estimate the average size of apples. If you simply averaged all of the fruit given, you might end up claiming they are about the size of a quarter. But if you look closer, some of those samples might have larger significance and be more useful than the others.

That doesn't necessarily mean the site is invalid or its conclusions, and I don't personally care much which "side" is right. But does that analogy help explain why this can be a tricky problem, and why many here, including myself, defer to "experts" like the NIH (who ran one of the largest studies of hcq)?