Personally, I prefer cognitive function theory because it gives a deeper reason behind splitting dichotomies.
Essentially, it suggests there are 8 major thinking patterns humans engage – which can be broken down into two groups (perceiving and judging) of two functions (sensing/intuiting; feeling/thinking) and two directions (inward or outward) - which is why we end up with a 4-letter code (and that could really be reduced to a 3-letter code, but regardless -)
I used to struggle with the why behind the theory not allowing something like high Ni and Ti, but theoretically, it's not about strength of function, but also role of function. So even if your Ti is second-strongest, it might not necessarily be playing the auxiliary role to your Ni. And those roles are described according to balance, which is a fundamental principle to the mechanisms of the human body.
The primary function informs your main outlook on life and provides a certain degree of motivation; the auxiliary function "guides" the first in curbing its information intake if the primary is Px, or broadening if it's Jx; the tertiary provides relief to the primary and allows an outlet; the aspirational provides support and balance to the primary; and so on - the shadow processes (5,6,7,8) perform similar roles but further into the subconscious, and tend to emerge on a more ad-hoc basis. And then that's why loops can kind of mess you up, because your functions aren't balancing out correctly. Generally the middle function between whichever two are looping is supposed to be intermediating, but is lacking. And then you end up either a hermit or really obnoxious.
But like [MENTION=5578]bologna[/MENTION] said, often JCF is misinterpreted and often tests are flawed. To be fair, it's very hard to test for functions. Just like with the Enneagram, you have to have a decent idea of what you're trying to identify and why, otherwise you get happy-go-lucky ESFJs who test FeNe and broody ENFPs who test FiNi and other such misunderstandings.