>>14339376>appears very sceptical of the validity of dichotomiesThis is true, and it's a problem. Authors of HEXACO used the same exact analysis that Big Five does, only they found six "big" factors underpinning personality. Big Five authors critique this as well as having too many lexical categories. In short, they're skeptical of anything that isn't Big Five.
https://psychology.stackexchange.com/questions/22078/what-are-some-criticisms-of-hexaco-from-the-big-five-campThe objections McCrae and Costa are raising to MBTI in that paper, and the page you show, is regarding the Jungian approach via "Extraverted Thinkers", "Introverted Sensors", etc. Sometimes this is branded as cognitive functions, and even then extended via Grantian function stacks. But none of this is how MBTI types are determined, and the 145 forced-choice questionnaire (form Q) indeed uses the facets as defined McCrae and Costa. That is to say, the point of the paper was for the authors to reinterpret MBTI in a way that's not Jungian, and rather a lexical way via facets, and then get meaningful results.
That's why they say ENTJs and INTJs differ from each other only by extaverts and introverts differ from each other. This is on contrast to the (pseudo-)Jungian approach whereby ENTJs and INTJs have the same functions, just in a different order.