The _threat_ of a polygraph test might filter out some guilty people who are afraid of it. They don't use the result of the test for anything, according to the article.

I think that ethically that's problematic. I consider that to be coercion based on pseudoscience. There are people out there who are afraid of it for a variety of reasons - I can think of a few off the top of my head:

They are concerned they might ask a question unrelated to the case that incriminates them

They are concerned that they ask a question that causes them personal problems, even if they are innocent (like past infidelity from 20 years ago, etc)

They are concerned that the polygraph might be wrong and incriminate them for something they haven't done

If they don't use the results of the test for anything, then it's a deceptive interrogation technique and one that should be stopped.