I could see both, and likely both would need to be part of estimation. It's not binary, but likely, 20% due to mask lowering viral exposure loads, 50% due to more healthy people (hence asymptomatic) getting the virus.

However, that doesn't invalidate the overall comparisons. Countries like Sweden have a lowering death rate despite no broad mask wearing (https://fortune.com/2020/07/29/no-point-in-wearing-mask-swed...). You can see an example graph here: https://www.independent.co.uk/news/world/europe/coronavirus-...