Look, the reality is that almost all interviewing is far from scientific. I'm sure there are worse heuristics many of us use than MBTI[1]. The comment being downvoted is being perceived as trying to appeal to an argument of extremes: It doesn't follow that if you're following some poor methodology that one should consider even crazier ideas.
If we had nailed interviewing down to a science, I could see the point in the comment. Instead it's coming off as "Hey if you're going to do something imperfect, you should consider something extremely wrong!" We all have mostly wrong approaches to interviewing.
[1] Stuff I've seen:
Rejecting a candidate because they read Chapters 1-4 but not 5 of a textbook when the interview prep material mentioned chapters 1-5. Note that chapter 5 is not used at all for the job.
Favoring a candidate because of a strong handshake (cliched, but happens!)
Gauging the enthusiasm a candidate by assuming everyone is an extrovert (fairly common).
