In any case, the large number of NDs (not detected) everywhere in the table, but especially inside the masks after the patient just coughed in them (which the authors struggle to explain), makes this entire experiment very suspect. (Just look at the reported results for patient 4, which make no sense at all!)
There is no such thing as a symbiotic virus.
That is when behavior comes in. If people are more likely to ignore social distancing if they wear a mask then you might get the opposite effect, as social distancing is seen as better protective measure than masks.
This is one of the reasons why the health department in my nation only recommend masks for risk groups, like if you are working with elderly people.
I recall reading some interesting research on this when it comes to playgrounds: these days unforgiving concrete is being replaced with bouncy rubber, and the compensation/homeostasis theory is that kids are playing in riskier ways because small tumbles don't hurt them any more.
It is not a black and white issue of seat belt where there is ample proven evidence that in basically any car crash a seat belt is better than no seat belt. There is no shortage of seat belts, nor has there ever been one. The focus of car safety has also primarily focused on speed, where the survivable limit goes up if the driver has a seat belt, but even that higher limit is still much lower than the speed of many drivers. Thus the message there is to inform drivers the importance of both the benefit of seat belts, but also the critical message that speed more or less determine if a person is going to live or die. Even if people would risk compensate and go from the survivable speed without seat belt and increase their speed to survivable speed with seat belt, it would still be lower than what many people already drive their cars at. The worst case scenario of seat belts would thus be an improvement. We can not say the same thing about social distancing, where going out sick with a mask on is still a very bad idea.
Thus the recommendation from the health department is right now that masks should go to those in risk groups and people who are sick should stay home. If research later show that people without symptoms reduces risk of infecting others by wearing a medical mask, and the supply of masks increases, then they will likely change their recommendation based on the changed situation and facts.
(Small meta comment, but it is interesting how skeptical people are of the medical department that is responsible for handling the pandemic. It is true that critical questioning any government recommendation is useful, but maybe a bit more open minded discussion would be better).
But they also say this: "The median viral loads after coughs without a mask, with a surgical mask, and with a cotton mask were 2.56 log copies/mL, 2.42 log copies/mL, and 1.85 log copies/mL, respectively."
That seems like a significant reduction in transmitted viral load? I would think that masks can have a positive impact even if they aren't perfect filters. If you look at their table of data, the reduction trend is consistent across their tests. But between the small number of patients(4) and samples, and the number of Petri dishes that came back as 'Not Detected', it seems like a pretty low powered study. :-/
They are also explicitly testing coughs. I wonder about the effects on filtering droplets generated from normal breathing and talking.