(1) It only takes one non-masked, close-contact exposure to coronavirus to become infected... that it, mostly. how it is spreading. The study does not control for the exposure risk profile of the sample populations so... there you go. This study is next to worthless...
(2) There have already been numerous studies, recently, that cloth masks are effective at filtering out particles, even down to 1-5 microns in some cases. Most cloth face masks are very effective and filtering larger particles (50-250 microns). Particle mass/volume scale with the particle diameter to the third power (i.e. V=1.33PIR^3 or 0.166PID^3). So a 100 micron particle contains 1000 times more mass/volume or viral load than an 10 micron particle and a million times more than a 1 micron particle. Thus is most important to filter the big particles in reducing viral load expelled by infected individual and breathed in by others. Cloth face masks, and surgical masks are pretty good at doing this. N95 and KN95 are already well tested in this area. So while masks can't filter out all virus... they are pretty good a significantly reducing the viral load which is important.
(3) I know people at work who became infected at a party and came to work for days while they were unknowingly infected. But thus far, there have been no "at work" coronavirus transmission cases because we wear good masks. Thus far this is true at our other sites as well. I know a local state epidemiologist who remarked to me that the overall rate of transmission in the workplace, for places with good mask protocols, is remarkably low.