(Polygraphs are pseudoscientific BS in the first place, but I know there are some cases where you must undergo them for whatever reason.)
EDIT: I'm not aware of any polys leading to discovering enemies of the state. Its all self-elected information from my understanding, in line with your suggestion.
Polys are not the end-all for detecting bad guys but if you choose to ignore it anyway, it certainly becomes useless.
I hadn't heard of the Ames poly history but I just looked it up and its interesting because apparently he had the same mindset that I do about polys in general [1], and basically when the poly tech said during the interview that they were getting some weird results, he essentially told them the same thing [2] and they were like "yeah youre right" lol.
[1] - https://sgp.fas.org/othergov/polygraph/ames.html
[2] - https://www.washingtonpost.com/archive/opinions/2002/04/16/failure-of-the-polygraph/07c406a5-0aa3-4e20-8dcc-89781162aaa8/In his original 1986 poly, he popped on one question but basically smooth talked his way out of it so it never was something that required followup (i.e. they were sufficiently convinced he was a good guy). In the 1991 poly it gets more interesting because the entire poly was partially a ruse because he was _already under investigation_. Due to timing, the interviewer didn't get coached to dig in on specifics around financial issues, and so they followed a bog-standard interview which Ames passed. The really interesting quote around this is how the postmortum basically says this was a bad poly because they didnt play enough mind games to get him nervous in the first place:
In its review of the Ames polygraphs, the CIA IG report quotes several current and former polygraph examiners who stated that the Ames case should not be considered to be a polygraph "chart interpretation" problem. Rather, they say, the fundamental problem is that the 1986 and 1991 polygraph charts were invalid because the examiner in each case failed to establish a proper psychological atmosphere in the examination sessions. A former polygrapher noted that without proper preparation, a subject has no fear of detection and, without fear of detection, the subject will not necessarily demonstrate the proper physiological response. Consequently, they surmise, the Ames polygraph tests were invalid because the process was flawed by examiners who had not establish the proper psychological mind set in Ames because they were overly friendly. As a result, Ames's physiological reactions were unreliable.