FYI: we're about to learn that Remdesivir "doesn't work" either. The Trump administration bought half a million doses of it in case it does work. So now watch for the "studies" to come out that say you'd be better off dying.
That site is rather... biased. And for all of its pretense of being scientific, the page with country data fails to provide usable sources for its graphs (the tweet doesn't seem to exist anymore).
Also, at https://c19study.com/boulware.html, a study where the authors themselves say that they found to prevention of COVID-19 cases, the site tries very hard to spin the paper into positive results nonetheless. Some of the numbers cited, especially the primary claim that "COVID cases are reduced by [49%, 29%, 16%] respectively when taken within [2+, 3+, 4+] days of exposure.", are not sourced or explained properly. I can't see any data in the paper regarding the effects of interventions taken within 2 days of exposure, so I find it hard to believe that this is somehow derived from the paper's data and the authors just missed it.
I didn't look at all the other studies, but if most of the "positive" ones are similar, then I trust HCQ and its supporters even less than I did before.