The recent mask study in Bangladesh comes to mind. I hazily remember it finding mask effectiveness at about 10-15%, which caused the press to proclaim that mask

The recent mask study in Bangladesh comes to mind. I hazily remember it finding mask effectiveness at about 10-15%, which caused the press to proclaim that masks work. However against an exponentially growing virus this reduction amounts to a brief delay in transmission. Happy to be corrected by someone who has looked at that study more deeply.