That's certainly a fine position to take but it doesn't change the results of the study: 40% of complainants admitted they lied. Unless you think people were lying about having lied, and also making up an explanation for why they lied, just because of polygraphs?

edit: I should note something else here. Kanin's study has an appendix where he reports a similar analysis on two university campuses. In this case polygraphs were not used and were thus not a factor. They also used the same definition of false (only the accuser can render an accusation false). The false reporting rate in both these cases was 50% - higher!