I think first link really oversells the study it references, though.Key quotes, from the paper: "The factor by which β was reduced was conservatively set to 2.

I think first link really oversells the study it references, though.Key quotes, from the paper: "The factor by which β was reduced was conservatively set to 2." "Varying degrees of mask effectiveness are modelled by the mask transmission rate T and mask absorption rate A, which denote the proportion of viruses that are stopped by the mask during exhaling (transmission) versus inhaling (absorption), respectively. Weset T = 0.7 and A = 0.7 to model the use of inexpensive, widely available, and even nonmedical or homemade masks [...]" So basically, they assume certain values for how effective masks are, and then simulates the dynamics of the epidemic. Nowhere do they try to justify the specific values chosen. I expect that the conclusions cited (if only 40% wear masks, there is little benefit; if 80% wear masks, total cases go down by a factor 1/12) to be heavily dependent on those assumptions, and so I would be cautious to read too much into it. Link to paper: https://arxiv.org/pdf/2004.13553.pdf https://arxiv.org/pdf/2004.13553.pdf