The claim is that, by the evidence, moths had ultrasonic hearing 28 million years before bats' own echolocation developed; and that some moths in environments without bats have since adapted to lower ranges. We must presume paleontologists can recognize the frequency range sensitivity of these structures from their preserved physical shape.
In what way would hearing not sensitive to frequencies that appear in the environment be adaptive? The existence of hearing adapted to ultrasonics implies the existence of something to produce ultrasonics. It doesn't tell you what those producers were, just that detecting them was important to successful moth reproduction.
One possibility would be the moths themselves making sounds. But why would they lose such a capacity?
It is possible they had to evolve to stop making the sounds to avoid attracting the interest of bats, but thousands of species all dropping it, even in places where there are no bats, is not plausible.
The best alternative consistent with what we know is that the original insectivorous producers of ultrasonic noise are now extinct, and that bats later evolved into a similar niche.
There are other possibilities that must be eliminated, through careful research, before we can be confident of late-Cretaceous echolocating insectivores.
