For example you might lose 10% of your salary but four people in India double their salary. It's a slow process so it will feel like the downward trend will go on forever but one day it will change its course and salaries will go up again.
Inequality within countries is increasing, and is more directly connected to experienced disutility (though media saturation from one country to another can also make inequality b/w coubtries a factor in experienced disutility.)
A problem with naive economic analysis is that it tends to pretend that absolute material wealth is the prime determinant of experienced utility (and consequently that disutility is mostly driven by absolute deprivation), which while true in the most abject poverty isn't true beyond that level.
It's not zero sum, but the article doesn't properly demonstrate that the positives outweigh the negatives. There is a wide spectrum between "total wages are fixed" and "total wages scale 1:1 with immigration", and the tradeoff is only worth it above a certain point.
I see the part where they linked a study about what was "often" the case between 1850 and 1920. None of the other links seem to be relevant to that part of the argument.
The article itself has a somewhat supported argument that it's not zero-sum. But it acts like disproving zero-sum is enough, when it really isn't. If bringing in 10 workers only adds 1 job, or 1 job's wages, that's not zero sum but it's almost as bad. And the article doesn't address how far it is from zero-sum, or link anything that does so.
coronavirus is already making remote work a first class citizen and removes the main advantage of a local workforce - physical presence (similar to how the global financial and shipping infrastructure removed the physical presence advantage of the local blue collar workforce back then), and the immigration ban is just a push for companies to fill more of those jobs with the foreigners in foreign lands where they are much cheaper. Yet the lessons of history seem to be ignored as usually.