I don't think he's arguing about your personal insight regarding your own relationship experience as an ENTJ who fits well with a certain INFP; he's arguing that you are misrepresenting this experience as representative of some significant majority of possible male INFP+female ENTJ pairings.
But you know what? If we're not allowed to do things like that, we may as well trade in the Ni for something that actually works. (Oh wait, Ni does work.)
Okey Dokey, now to pretend that Ni has a basis for all the massive leaps from particular to general it does, particularly on the question of type relations:
Assuming all people do have a type, people make of their type whatever they make of it, but still it's there. The details of the interactions between two people of a given type will vary according to education, habit, baggage, culture, location and a billion other variables but... if type is real, then a foundation does (or in mismatched types, doesn't) exist for adequate, quick, deep communication.
What could stand in the way of that foundation? Well, type theory is a moderately expressive model of cognitive goings on. It appears more or less scientifically unprovable, everyday experience not being a proof that the model is anything other than moderately descriptive. So, there could easily be some big fat psychological process sitting there, hidden away behind layers of other mumbo jumbo we use to understand ourselves, and that hidden, undescribed, non-MBTI modelled process could well be fucking with what we might call MBTI-stuff.
In other words, MBTI is moderately descriptive, but cannot be called complete or truly deep. There could be, maybe even probably is, other stuff at work. And that may make any type claim invalid.
Or not.
Who knows?
INFPs and ENTJs go well together.
And on second thoughts:
Hm, nothing proven except that some people, me included, appear to like to generalise. But how could Te focused Ni be wrong? Well, it's not ever going to be right unless it has experience of the world out there. And it's not ever going to have complete experience. And is it ever going to be able to prove that it has nonetheless had adequate experience?
It might. Ni/Te (if it exists and functions the way it's described as functioning, and doesn't suffer other mysterious impediment) will learn to generlaise adequately, and be satisfied that it has done because it's not Ti. And Because it's not alone: there's Se as well. The Se person is going to cruise along soaking up new stuff even as the Ni/Te person makes conclusions.
Put Ni/Te in a box and shut off its access to the world, and it'll be demonstratbly worng a lot!
Ah, the joy of a judging function... you don't have to be endlessly engaged in gathering information without end. And the sadness of a judging function: if you don't endlessly gather information, you'll always be wrong.
Why, it's almost as if there has to be a balance. Luckily, people have different focuses, different tendencies, indeed, different preferences for how to process. By jingo, I'm right because I'm right because I'm me: if there weren't other people of suitable type to balance my type, I wouldn't be able to go so far into my type, and be me: I'd have to be Everyman instead.
So where do types come from? I wasn't born with the knowledge that someone else can balance out my extremes letting me go to my extremes. Or was I?
This last question seems to crystalise the inadequacy of type theory. It's tendentious. It doesn't explain. And when thinking it over, one comes to the expressive limit of MBTI quite quickly and knows there must be some better, deeper explanation.
But practically speaking, INFPs and ENTJs go well together.
And it's kinda sad that that's the best I can say about it. I don't know why it's sad. But it is. Were I a believer, I'd say my Fi is moaning. Which seems to be true. But we know it isn't, because MBTi is just a model.
