Good questions. There's two things to look at - population level evidence like your discussing - then the laboratory side of why masking might work. I'm not super up to speed on the answer to your question. I know those studies exist but I haven't done a deep dive. What I've briefly reviewed has all found benefit to masking.
My early support of masking (had my mom start sewing home-made masks early March) was based on limited data and inferences that have held up. I found this surprisingly great article early on (
https://smartairfilters.com/en/blog/best-materials-make-diy-masks-virus/) that linked to good information. The basic takeaway was that crappy homemade single layer masks could still capture greater than 50% of viral particles smaller than COVID.
Thus a super simple premise is born. IF infection is spread via airborn micro-droplets, crappy masks likely reduce the exhaled viral load by 50%. If a single viral particle leads to an infection, then masks probably do very little. But if infection is dependent on viral load/time exposed, then masking would logically lengthen the time needed for an infection and reduce the viral load in an enclosed space.
Cost/benefit becomes a no brainer. Even if the impact is very small, the cost benefit is there. Particularly when you consider the nature of viral growth. If all else constant, mask wearing only reduced infection rates by 10%, that would result in roughly 1/2 the total case count after 6 generations.
If you combine that with lab videos of talking/coughing particle transfer with and without masks, you have a really good logical argument to masks being beneficial. Thus, I think the burden flips here. The risk/reward or cost/benefit strongly favor masking. If you were wrong and they didn't help, the cost was negligible. But if they do help and you're wrong the other way, the cost comes from increased hospital stays, potential business or school closures, etc. So there's an asymmetric risk/reward here that strongly favors masking.
Study with hamsters - using a surgical mask barrier between cages reduced infection by ~60% (over a week I think) AND reduced severity of infections. There's LOTS of evidence that viral load matters here.
Lots and lots and lots of evidence for masking, almost a total dearth of evidence against it. Most of the evidence against it is pre-covid (based on flu or other viruses).