Yes, for the following reasons. At the moment an infected worker comes into proximity of a resident, the air stream will deflect his infectious aerosols away from the resident during the interaction. It is at this moment and the few minutes that follows after the start of an interaction, that protection against infection is the highest, while in contrast without the solution, protection would be at its lowest.
That initial protection from the crosswind unit could even be better than 1 infection in 10000 interactions (compared to the relative baseline of 1 infection in 1 interaction without the solution). Because it takes 20 minutes to circulate the room’s volume though the crosswind unit once, during middle of that period the protection will drop from 1 in 10000+ to 1 in 1500 short-term improvement, the worst case in the time progression for protection. But that short-term worst case on its own is already a huge improvement over no solution.
Therefore, because the crosswind unit is continuously clearing out the infectious aerosols, after 10 minutes of initial contact, the "1 in 1500" statistic begins to improve. Let us say the interaction was 5 minutes and the worker left the room. So after 20 minutes, the probability of infection drops to 1 in 3000, and after 40 minutes, it drops further to 1 infection in 6000 interactions, and so on (always comparing to 1 guaranteed infection in 1 interaction without the solution). That means if the resident did not get infected in the first hour, the chances of her getting infected are lower than 1 in 6000 after one hour following the departure of an infected employee. If the infected employee does not return, after 7 hours the air purity in the room will reset back HEPA filter quality, and will maintain HEPA filter quality if no infected worker enters.
Because no one knows when an infected worker can enter a room only find out days later from screening tests that he was infected, the crosswind unit has to operate 24/7.
To understand the statistics better, consider the converse scenario, with no crosswind solution. Let’s say that the resident was lucky and she did not get infected by a masked but infectious employee in the first 20 minutes after his departure. Then his infectious aerosols would diffuse into the room volume and remain suspended in the air, available for infecting the resident at a probability of 1 in 1500, not only for the first 20 minutes, but also after many hours the probability would remain the same instead of continuously reducing [view here in Part 3]. If you compare the crosswind solution to a surgical mask, the mask by itself reduces aerosols by 90% emitted by the infected worker [view here in Part 3]. The crosswind solution further reduces infected aerosols emitted from his mask by up to 99.93% after the first 20 minutes of operation (see Question Q18 to find out where this statistic comes from).
The improvement is dramatic after only 20 minutes and continues to improve with time. The unit only consumes 30W of power, so it is very energy-frugal. Under normal conditions, that means any time a worker enters a room, he walks into an environment of the purest air possible. Please keep in mind that these statistics do not directly model the probability of infection. Instead, they model the improvement in the reduction of the probability of infection, if this solution is applied on top of current mitigation measures. For example, in October 2020, this region had 40 long term care centers infected with 320 cases due to the 2nd wave:
https://www.cbc.ca/news/canada/montreal/christian-dube-chslds-covid-19-1.5773049
If this solution had already been in place in September 2020 in all the centers in my region, there would have been no infections due to any kind of airborne transmission.