Basically all convincing evidence for mask use is models (like the linked article) and in vitro unrealistic tests; whereas in real world data, the best I could

Basically all convincing evidence for mask use is models (like the linked article) and in vitro unrealistic tests; whereas in real world data, the best I could find wrt covid is 2% reduction. Bacteria are a different matter, and possibly other viruses. But real world data about covid is that masks make very little difference either way, and are possibly harmful.