What would you propose we do instead? Give a group masks and tell a control group explicitly to not wear masks, then see who contracts COVID more often? There

What would you propose we do instead? Give a group masks and tell a control group explicitly to not wear masks, then see who contracts COVID more often? There is such a thing called ethics. You can't experiment on people when you have good reason to expect something will be harmful for them, like feed them cancerous stuff to see what happens. But you can check what happened to the people that didn't comply, for example those that accidentally ate that food, and still learn something. Telling people to wear masks, even if there is a chance it doesn't help, is not a weird idea during a pandemic. The chance of it being counterproductive was fairly negligible compared to the chance of it helping, and so you should do it. Or did I misunderstand what you meant there?