No, the safety data is not well controlled in this case. There are lots of ways to administer a vaccine, it doesn't have to be injection.

Look up the history of vaccination. Starting in the late 18th Century, people used to do this exact thing for early vaccinations: they found a naturally occurring, less pathogenic strain (for instance, cowpox[1] in the days of smallpox), and intentionally exposed people. This is exactly the same thing as a vaccine except less controlled and, ultimately, much less safe.

.

[1]The root of "vaccine" is "vacca," i.e. Latin for "cow," for this reason.

Safety data can be gained from epidemiology studies. There is more than one way to work out if a treatment is safe and effective.

Epidemiological studies are much poorer than clinical trials. I’d rather have low-N but well controlled clinical trials than nebulous epidemiological studies.

Of course, but in this pandemic we shouldn’t let the perfect be the enemy of the good. Waiting for the perfect vaccine to be developed will have a very high cost. There are no easy options here.