Ok I've read the actual study and found that the most relevant table is in a different section. It shows that surgical masks were indeed effective for all age groups but cloth masks results were inconclusive. Edited your answer to provide a better overview of the paper and removed my skeptical comment about the 18-50 age group. IMO this is now definitely the best answer.

This looks promising but I think the table needs more interpretation and some conclusions, i.e., what does it mean for an "intervention coefficient" to be negative? Does the absolute value tell us anything? Is the "baseline control" the mask wearing instruction, or something else?

1st ref is theoretical BS, 2nd ref from Bangla is practical BS: a total fail on all levels, and contrary to what's asked here, for personal protection, allegedly. Please note that if the study were to be taken seriously—and it is in no way on many levels— it would show that masks do not protect others from infected ppl spreading, it, and that again for the arbitrary def of age cohorts, it failed —utterly— to protect the younger generations -> absolutely useless in either direction for ppl <50yrs (fig3, p28)! One can be confident that using this study to say 'masks work' is misleading.

@JonathanReez You awarded a big bounty on an answer that fails to bring up convincing evidence—and you knew that immediately as displayed in your first and now deleted comment which was asking to the point about the plainly ridiculous age anomalies? You even edited the relevant tgraph out of this A! This looks quite manipulative in 'reading' the data presented.

There's skepticism, which is often useful. As evidence is uncovered, skeptics evaluate it and change opinions. Then there's what you're doing. If you can be entirely replaced with a piece of paper that says "If I don't believe it already all evidence is BS" why are you even here?

Well, can you explain 1. how that 1st study addresses the question asked (efficacy was excluded, so there goes the theoretical first one ref) and how the 2nd can be evaluated in terms of proper planning and conduct, that is about PPE not src-control, with a graph you chose (initially) that clearly fails to demonstrate any relevant effect for mask wearing in ppl below a certain age, with even a very small effect in such a gargantuan sample (7 in 10000) extrapolated to all? It doesn't show efficacy, but terrible design and pompous claims for reach without much internal or external validity.

No. Your comment is word salad and you are not in any way acting in good faith. You are not a skeptic, you are a dishonest pseudointellectual and there's no use engaging with you at all. Downvote and move on. Your schtick is tired.

@LаngLаngС if I didn't award the bounty it would go to the most upvoted answer on this page automatically - and I don't like it

the data does show some effect from mask mandates based on a large real-world study. Its as best as we can get short of an actual mask challenge trial

btw personally I'm strongly opposed to mask mandates (Now that I'm vaccinated)

so if your attack was focused on calling me out as a feverish masker, this judgement was misplaced