^ yes. O_O;
i don't have any problem with the idea of an additional letter, but i don't really find it necessary, either. part of the whole reason the system works is because it does force you to choose sides, even though you may in truth be very very close to the middle. but that's true for everyone... some ENFPs are ENfp, some EnfP, some eNfp, etc. like others have pointed out, we could add many more letters, but the system is a fairly balanced set of dichotomies as it stands.
personally i just tend to think of it as cognitive, rather than social, extraversion, and it makes things much easier. i also have always scored I in tests but upon learning the functions, identified as a Ne-dom almost immediately.
anyway in my mind it works like this:
EJ : Extraverted ; Judgment extraverted : JePi
emphasis on prioritizing info outside the mind ; secondary emphasis on collecting info inside the mindEP : Extraverted ; Perceiving extraverted : PeJi
emphasis on collecting info outside the mind ; secondary emphasis on prioritizing info inside the mindIJ : Introverted ; Judging extraverted : PiJe
emphasis on collecting info inside the mind ; secondary emphasis on prioritizing info outside the mindIP : Introverted ; Perceiving extraverted : JiPe
emphasis on prioritizing info inside the mind ; secondary emphasis on collecting info outside the mind
i think part of the problem that we run into is that extraverted Perceiving is a bit of an odd notion, because it is extraverted, but Perceiving is really a somewhat passive activity. (no wonder we can seem lazy...!)
so that helps see why an EP might seem less extraverted than an IJ, because EPs don't tend to prioritize externally. unlike Je types we don't generally seek to prioritize our environments, even though we are focused on gathering information from them. so we may tend to act outwardly less, even though we do tend to express outwardly. whereas an IJ with an external agenda is going to seem more extraverted. and then of course having S, especially Se, adds a grounded depth that can make those types seem more classically extraverted than their N partners as well. T, with its rooting in objectivity, also seems to lend more groundedness than F.
we are also going to retain some of the "contradictory" characteristics of our secondary process - i have also noted the "time away" that ENxJs seem to need, and i very much suspect that is characteristic of Ni processing.