When I first discovered MBTI, I read and reread these different type descriptions obsessively until I could quote them. After that, I was able to simply tell people (whom I knew well) which type they were, because in my head I could picture the person next to the entire article, line breaks and all, and kinda jam them together and see how well they fit. This was not a rational method. If I could put it in written form, it would probably look like a fisk of the closest MBTI archetype description (minus the functional analysis), full of disclaimers, complete with a conclusion that I’d be entirely unable to defend in even a pre-scientific way. My apologia would sound tautological: “You’re this one ‘cuz it feels closest.”
It also didn’t work very well for typing new acquaintances. For that, I developed a rational method of step-by-step functional analysis, which was more difficult and slightly less accurate (still pretty good though). This would have lent itself quite well to writing structured descriptions of people, whose outline would look something like this:
I. Thesis: Person X is INTP
a. General introverted traits
b. Introverted thinking
c. Introverted sensing
The lower levels would just be checklists of the traits within the aegis of the subheading. Here’s an example list of traits I could use for subheading IIa (it’s a little cynical; quite rightly, sez I).