So according to this model, best use functions in the INTJ example are Ni (dominant), Te (auxiliary), Fe, and Si (crows nest). It's the tertiary, inferior, and double agents that tend to cause the most problems.
Well, that ordering of function development looks a little different than what Andy suggested, unless I'm missing something. It does make sense to me that 3 comes before the others. That being said, with all the Tertiary "trip wires" that seem to be out there, it seems to be one of the more troublesome ones in any scenario.
Maybe from a practical perspective, a combination of Naomi Quenk's Eruptions of the Inferior (Option 1 described above) and Lenore Thompson's discussion on Double Agents (and maybe Crow's nest) might be the best way to think about some of this stuff. I'm not quite so familiar with Beebe's work and how it fits with those things.
It's supposed to be the NTPs who worry about the precise usage of words, but to explain how I view the types and type development I think a quick lexicon is in order. Note that these are just how I use the words, not necessarilly everyone in the world.
Function Order: This is the usually 1 - 8 list that gets seen and is concerned with how the functions are used, when they are used.
Function Preferance: How much you actually use each function.
Function strength: How easy it is to use a particular function in a healthy manner, rather than it misfireing and giving you problems.
Function development: Whether or not a given function does work well for a particular person.
To my way of thinking function order and function strength are invarient for each of the 16 types. They are the same for all individuals of that type. However, function preferance and function development can be radically different from person to person, even in the same type, which creates the differances between seen between people of the same type. It also explains many of the differances seen in the development of the types.
Thus, I don't believe that function development has to follow the same pattern for each individual at all. Some people may find that the critical function gives them better results than the teriary, while for others the reverse may be true.
If you like, function strength is like a statistical average along the lines of "At age 21, 50% of ISFPs will get good results from their tertiary Ni."
To me, the debate over function development is really about function strength, those averages that are so hard to get a grip on. It's hard to say how many members of a type at a particular age will be able to use a function well because it is hard data to obtain, especially as growing into a function is like having your voice break. It will flit back and forth between working well and malfunctioning for awhile. Even in old age, a poor function may still cause trouble on occation.
THis is why I prefer a Beebe style 1 - 8 ordering, with the concious functions on one side and the shadow functions on the other. You see, while I'm sure that going from 1 through 4 follows decreasing strength and 5 through 8 show increasing strength, I'm not convinced I know the relative strengths of the concouse functions compared to the shadow functions.
Instead, I simply except that enough fuzzyness, (enough standard deviation, if you like) exists in the function development that variations will exist in the actual order seen. Thus if asked which will develop first, the 3rd or the 7th function, I would tend to reply "wait and see." I don't think it is something that can necessarilly be predicted ahead of time with ease.