It seems likely that when you get infected by the coronavirus, the quantity of virus that infects you makes a big difference in how seriously ill you get. The logical (unverified to be sure, like most of what I will say in this post) explanation for this is that, if you get only a small dose, your immune system (assuming, of course, that it is healthy) gets a head start before the viral load builds up and you get seriously sick. It seems to me that the number of medical personnel who are getting seriously ill and even dying is evidence (not proof, not yet) for this effect since they might be getting huge doses of virus.
Hold that thought.
Variolation was a technique used before smallpox vaccine existed. People would take a sample from a pock and scratch their own skin or that of their children. It was dangerous and some people died. But you would mostly get a mild case of the disease with only a few pox and thereafter be immune. Needless to say, this was highly controversial and with good reason. But overall, it saved a lot of lives and a lot of disfigurement. Part of the problem was that there was no good way to regulate the dose. Viruses could not be seen until the invention of the electron microscope in the 1930s.
I think you can see where this thread is heading. Nowadays, I expect that they could more readily create a standard sample. Would it make sense to try to inoculate people, especially medical personnel (and only a few at a time) with a tiny sample of the virus and let them have the disease, hopefully mild, and then let them have at treating it in others? I’ll bet you would find volunteers to test this. You can find volunteers for most anything. Think of this as a war. It seems less dangerous than being a spy in Germany during the war.