I was serious with that comment. You can do longitudinal studies now, and people do, comparing infection/hospitalisation/death rates among vaccinated and unvacc

I was serious with that comment. You can do longitudinal studies now, and people do, comparing infection/hospitalisation/death rates among vaccinated and unvaccinated people. But those have very serious confounders, namely that those who are still unvaccinated might behave rather different from everyone else. To overcome this, you'd need an RCT. But you're unlikely to find either a) unvaccinated people that want to be vaccinated, yet would be ok to get a placebo instead, or b) unvaccinated people that don't want to be vaccinated yet would be ok to get a vaccination instead.