I wonder how long it will take for the mosquitoes to evolve a mechanism that allows the female to detect males infected with that bacterium and not mate with them, or a mechanism to counteract the detrimental effect on the development of the eggs.
Right now the evolutionary pressure to do this is probably quite low as there is only a small number of infected males, but if that number goes up the pressure will increase exponentially.
Example:
If 1 % of males are currently infected with the bacteria, the selective advantage of a mosquito that can counteract the infection is just 1.0/0.99 (as the probability of mating with an infected male is just 1 percent), which is probably too small compared to other risks to produce any visible evolutionary effects.
If we increase the number of infected males to 90 %, the evolutionary advantage of detecting them soars to 1.0/0.1 = 10! This means a mosquito able to detect or counteract an infection is ten times as likely to produce offspring, which provides an incredibly strong gradient for evolution.
The question is of course how fast an immunity can arise (or if it already exists in the population), and how many generations of mosquitoes are able to survive after the infected males are introduced.
Probably they ran their own population genetics simulations on this, so I'd be curious to see results, which should give a good indication on whether this can work and if so under which conditions.
My personal guess is that it won't be effective, as there are very few cases where introducing a single external stress factor into a population causes it to collapse entirely, what's more likely is that it will adapt and relapse.