A topic that has been coming up for me in different contexts with different people is focus versus purpose. I think some people are focus oriented. Specifically the Si dominants. Recall my theory of type is that the functions are fundamentally information streams in the cortex, thus manifest as motivations in the psyche. This is because if you can handle one type of information stream over another, you'll be motivated to seek it out.
Accurate, as far my observations of people's behavioural streams matches up to their type. Fe-doms make decisions based on how they feel. Ni-doms make decisions according to what matches their inner visions. Seems to work. Good way of describing what Jung was getting at. Also matches with his profession of psychoanalysis, being to resolve people's irrational behavioural patterns by understanding and resolving their inner thought patterns.
Si is a focus kind of function, as is all the introverted functions.
So I think the better way to think about it is that INTP's are Purpose driven.
Contradiction. Requires a resolution.
I contrast Se versus Si this way - a Se dominant sees a painting and takes it all in the first moment, but doesn't catch much of the detail. A Si user only looks at one thing at a time. The face (for example), the hand, what the hand is holding, travelling around the painting. Our memory systems respond well to this kind of focus which is why Si types have excellent memory for past events.
Good point. I've observed the same about ESxPs and ISxJs. However, I found it makes more sense, when I visualise both types looking at a painting in an art gallery. The Se-dom acts as if he's looking at the painting from the back of the room, seeing all, but very little in great detail, only if he focusses, and even then, he quickly has to switch to see something else. The Si-dom acts as if he's looking at the painting from up close, almost touching the painting. He can see the slightest flaws and even discerns the hues of the colours. But he doesn't see "the big picture", and only focusses on what he is looking at.
I see the same sorts of things in Ne-doms versus Ni-doms, except they are looking at people walking past the painting. The Ne-doms tend to pick up on the general patterns of who stops to look at the painting and who doesn't. If a traditional landscape, more conservative people tend to stop and look. If an impressionist work of art, then the more unconventional and bohemian types tend to stop and look. The Ni-dom tends to focus very intently on a particular individual who walks through the gallery. He studies the person, can even predict which paintings he will stop at. He makes generalisations about the rest of the guests on the basis of his observations of that one person. When he is talking about someone who is extremely like the person that he studied, he's eerily accurate. But the more different they are, the less accurate he is.
And so Si types tend to talk in terms of focus. You especially notice it with the sports Si types. They keep mentioning how focused they were (or weren't). But what are their goals? A lot of the time, not much. "If I make trials for the Olympics then good, if not that's OK too". Not a lot of purpose there.
They tend to be
project-oriented. When on a job, nothing can stop them from finishing their task. They'll work and work until it is done. If, however, the job ends, like they trained all they could for the Olympics, but didn't make the team, then they've finished the job, which was to train for the Olympics, and move on to the next task.
At a young age I discovered that when I had purpose, a long term goal, plan or idea, and when I internalized it, I felt the most myself. When I didn't have a purpose then all the loose ends led to mild anxiety, depression and listlessness. But conversely, being hyper focused on something didn't work either. It made me somebody I wasn't. So, while I'm always going somewhere (purpose), I don't focus too closely on how I get there. This seems to be optimal existence.
That's an accurate description of myself as well.
But I would hesitate about drawing any specific conclusions about it, unless I could see that it was a
general pattern of most INTPs (Ne acting there). If I don't have that general pattern, then my deductions and inferences tend to be hit-and-miss, 50% of the time dead-on, and 50% of the time completely out of touch with reality.
What about INTP's? Dominant Ti, introverted thinking, focused. Yet you can see, and I'd argue, that we're actually hybrids and should be thought of as dominant purpose people. The reason is our auxiliary, as much as it is bracketed by the focus Ti and Si, Ne is not focused. It is diffuse, scattered, jumping from one to another. This causes many an INTP stress as they want to be focused, and am, but Ne keeps upsetting the apple cart.
So I think the better way to think about it is that INTP's are Purpose driven.
Jung described the difference between extroverts and introverts, as the difference between the
prolific and the
devouring, respectively. The extroverts are
prolific. They have the way-off view, doing many things, but never focussed on any one task that much, valuing their overall value, and valuing each task by its average value. The introverts are
devouring, focussed, very intent on the particular task they are preoccupied with right now. They'll throw everything into it, but spend a lot of time on it, and don't worry about it their average productivity is much lower as a result.
What I have also realised, is that Te-doms, Fe-doms, Ti-doms and Fi-doms are all very
result-aware. They all seem to be very aware of how much their current efforts will be likely to achieve the goals that their tasks are intended to produce. This is in stark distinction to the Ne-doms, Se-doms, Si-doms and Ni-doms, who remain totally focussed on accomplishing their task, and seem very oblivious of when their current tasks are extremely unlikely to achieve the goals their current task is intended to produce. They only seem to concern themselves with the pointlessness of what they are currently doing, AFTER they've already finished the job. But they are very, very aware of what they are trying to do at the moment, and so are very
project-aware.
I'd be more inclined to say that IxxPs are very focussed on what they do, but are also very concerned with if their current task is going to achieve the goals of the task, which is often uncertain. This makes them quite unsure of if they should be continuing with their current task. It also makes them very aware of when things look like they might go south, and make the whole attempt worthless. They are also extremely motivated to figure out what might be done to ensure that the current task can be guaranteed to be achieve. Both of the latter factors make them excellent advisors to ExxJs in charge, which is one of the reasons why their opinions are so valued, even though they themselves often seem to think that their opinions aren't worth much consideration at all.
As for Ne-aux versus Se-aux, I've observed, that when ISTPs are unsure, they resort to what they know, or what is conventionally known to work. It usually does. It's not that efficient. But ISTPs just get on with things, and it seems to work for them.
INTPs seem to dither a lot. I suspect because of that nature of Ne seeing many different possible patterns.
Te-doms & Ti-doms both seem to keep their focus on the question of if the job will be achieved. When problems come up, Te-doms go for adding another task to the plan that will solve the problem. Ti-doms try to integrate the solution into their original plan. But with Ne seeing many possibilities, which one to pick that will be the most likely to achieve the result, and will be the most efficient, becomes quite difficult to measure, as several integrated, highly complex plans need to be drawn up and each measured against each other, when the nuances will be small and the differences may be only slight. Too much thinking needed, for a very small payoff. Hence the lack of focus over which plan to pursue, and the confusion over even trying to figure out which plan to pursue. The differences between the overall value of each plan may be slight, and so the whole task of figuring out which plan to pursue seems worthless. But with knowing several plans to pursue, and without knowing which one to pursue, one cannot move fowards. Paralysis analysis, the oft-repeated complaint of INTPs about themselves, of others about INTPs, and of your issue that you mentioned, that seemed to give rise to your answer.
A key that I have found is
versioning. As you've remarked before, and I heartily agree with, Ne in INTPs seems to flounder when there is no existing framework to build upon. Ne extends. It does not create ex nihilio. It needs an existing Si framework to build upon. Get too far into my thinking, and I become confused. The bosses that I've worked with, who got me to work well, made me make versions. When I did something, I had to complete the details, till it was complete as a feature update in itself, even a very small one. Then when I got confused, they'd make me go back to the previous version and repeat my work all over again. Seemed very laborious and pointless. But it has consistently worked to resolve my confusion, and get the job done to completion. I've also realised that when doing my own personal tasks, the same principle has served me very well and very consistently.
From a functions POV, Ne can build on Si. But if Ne sees 5 ways to move the next step, and then 5 ways to the step after, and 5 ways to the step after that, that's 125 steps in total. It's too much. So what I find much easier, is to focus on the next step alone, work out which step would be best, and then solidify the next step until it's reliable, i.e. developed until it is satisfactory to the standards of an Si-dom. Then when I move to the next step, I might see a quicker way to do the whole thing, but I can still keep moving forwards, without keeping on re-writing my code and never completing anything. Usually, by the time that I've got to the end of the project, I've learned enough, that I can add a new method that completes the whole task much, much easier than any of my original plans. If I have the time and there's benefit in doing so, I'll do it. But often, by that point, it's just not worth it for the current project. So I log it, and then apply it to the next project.
For the TL;DRs:
Versioning -> See the 5 ways to do the NEXT STEP, and nothing more. Pick one. Solidify it, until it's Si-satisfactory. Save it as a new version. Then move on to the next step. If I become confused, just go back to the previous version, and repeat the process.
Same for when I have no way to figure out the next step. Refer back to my standard general solutions. Try them all. One usually works. Then I have a solution to move forwards.
. 
