A) It is one study. Very likely not statistically relevant on its own.

B) It is a certain subset of the pricing world that was tested. It may not apply to yours.

C) They themselves say "we found remarkably little impact on revenue". They looked for something and didn't see it. It's like you sitting in the forrest trying to see a boar. Just because you didn't see one doesn't mean there is none. For instance, one reason they didn't see the expected results may be that their experiment is not the right one, or that the tested population has an exceptional variable they didn't know about.