> So if you're confident that HIV has similar levels of spread
It does if you do phase III trials in the right countries. In South Africa, there's 15 million HIV cases give or take, and a million infections per year - so trials could yield results rather quickly...
In practice the trial will recruit participants from only a few countries. For example Pfizer's trial of BNT162b2 (their COVID-19 vaccine) recruited in the United States, Germany, Turkey, South Africa, Brazil and Argentina. But not for example Spain, Canada, Israel or Zimbabwe (picking countries off the top of my head at random).
In the ideal case you'd somehow randomly sample the world population, but practical considerations have to be taken into account. As a result it is often the case that we don't yet have a medicine targeted precisely at your gender, age, ethnicity, build, pregnancy status, co-morbidities, profession, and favourite musical genre, but ultimately the medicine which worked on another human who is quite unlike you in almost every way is surprisingly likely to be effective anyway, and so maybe you should try that meanwhile.
It does if you do phase III trials in the right countries. In South Africa, there's 15 million HIV cases give or take, and a million infections per year - so trials could yield results rather quickly...