We have several subtype system, as you know. But maybe these don't take it far enough? There is the 2-4-8-16 subtype system and they're all linked. But why should it end after that? One major accusation socionists have to deal with is the fact that the theory can't encompass the human nature in all it's complexity. That's true, because it's just a mental abstraction of the huge amount of different personalities.
But if we think of the sociotype as a fractal, which goes on and on if you zoom in (like a mandelbrot set), we're one step closer the the real nature of personalities. This was also suggested before (probably even in jest) but maybe we should see every humans sociotype as a chain of the the known "main" types. That means you basically have all 16 types in a certain order, fitting your individual personality, your main, secondary tertiary type, ect.
What if you only behave like you idealised and standardized main type suggests about 70% of the time? You might doubt and change it, but actually you know that this one probably still fits best. But this could be the typical situation and not just an exception. If we assume that you act and feel like your main type about 70% of the time, all the other types left in the chain also have a certain percentage. Your second type maybe 7%, the next 3% and so on. Well, we have to stop after the 16th type, because after that it would just repeat itself, but the last types would have minimal importance anyway.
I actually still doubt the practical merit of the dual type theory, but theoretically, it definitely makes sense. And I came to the idea that this "second" type might influence your self as a whole to a certain extent. If I'm actually INTj-ISTp, then my dual seeking function of my main type would be my secondary type's polr. I'm seriously wondering why I believe that LII fits very well, except for that one point. The point in which INTj and ISTp extremely clash. I'm pretty sure I'm not the only one who thinks this way and maybe the answer can be found in your dual type.