> Why default to not try it?
That's because ivermectin is one of tens of thousands of substances we know are VERY EFFECTIVE for something, but we can't give them all to each patient. Even if all of them costed zero and had no side effects ever. They just wouldn't fit in the patients stomach or bloodstream at recommended dosage.
So we need to pick some substances. And it would be ideal if we picked based on something more than pure luck (hunch being correct is still just luck). So we actually need to measure how good any given substance is for covid. But it's not that easy, bacuse we don't have a good system for conducting randomized controlled trials quickly and in organized verifiable manner. So inital studies are just doctors trying something on few of their patients often without any statistical rigor. You can still publish this as a study. You just have to write some stuff down. It doesn't have to be all the stuff. You can take 'out of sight, out of mind approach' with patients that don't fit your hopes. It happenes all the time. And when you get no success there's not much for you as a doctor to publish.
Covid is hard to track because it's very survivable so most patients that you treat will survive regardless of what you are treating them with (if anything at all). So you may very easily fool yourself into thinking that you are helping.
That's why it's better to wait for large randomized medical studies done by medical researchers as impartial as possible. Because every medicine has some side effects at some dosage so the chance of getting any value out of random medicine is nearly zero and chance of inflicting harm when people will safe-medicate based on rumours is significant.
And even if it has zero side effects medicines fashionable in context of covid already have patients that they should be given to. The ones that suffer from all the things that we know those medicines are VERY EFFECTIVE for. So if you don't ramp up the production to give most likely non-effective medicine to people that most likely don't need it, you'll be stealing it from people (and horses!) that do.
You don't need 1000 studies. You need one that is large and good.
Why not take chloroquine? After all it doesn't hurt, right? Or amantidine, highly fashinable in Poland, because one doctor believes in it strongly and advocates for it loudly, although reporters found out his track record with it is not as good as he's saying. But what do they know, right? Or maybe we should inject blood plasma of covid survivors? Sounded reasonable, many doctors used it for treatment. Turned out it doesn't work. Or hydrocortisone, it's just a mild steroid that doctors use to treat severe covid with effect of at least few percent. Or budesonide, another steroid that I personally think they should be using instead because effect looks way stronger. Or why trials of Fluvoxamine are stuck? It was looking so perfect in few initial studies. It's actually my favorite potential covid miracle cure.
The fact that you know of one drug that might be doing something doesn't mean we should be trying it (except for controlled trials) because there are thousands of exactly as promising or more promising substances and we just can't try all of them haphazardly because of the suffering that would cause to patients that don't need the drug, and the patients that actually need it and won't be able to get it.