To put those in context, what the study did was promote masks across a bunch of villages. Some people across all arms of the study were already wearing masks of

To put those in context, what the study did was promote masks across a bunch of villages. Some people across all arms of the study were already wearing masks of various sorts, so the data should be read as the combined effect of: the promotion changing behavior, people starting to use masks or replacing their existing ones, and then the net effect on transmission. So the % effectiveness quoted is the effectiveness of the new masks relative to whatever people were already doing watered down by how much putting up posters etc actually changes human behavior. Pretty remarkable actually.