Earlier quoted context omitted.
When 70% of people weren't affected by the experiment (13% were wearing masks before the experiment started. 60% of people didn't wear a mask after the experiment), then of course the numbers are going to be small. Seeing a measurable difference when the population only changed from 13% mask to 40%ish mask usage is pretty outstanding IMO.
I don't think that's correct; see https://news.ycombinator.com/item?id=28407881 for the argument in more detail.
That means that the full matrix of situations are available: I'm correct and you're correct. I'm correct and you're wrong. I'm wrong and you're correct, and finally I'm wrong and you're wrong.
----------
So lets just try again from the top. The study shows that an 8-week mask-awareness campaign in Bangladesh across 300 villages (+300 "control" villages 2+km away where mask-awareness did NOT happen) leads to a 29% increase in mask usage and an 11% decrease in COVID19 cases.
Is there anything wrong with this new, revised statement I made? It seems to be the big "A / B" test that everyone's wanted for mask studies.