Ne and Ni in Feeler Types

Sorry if this has been done to death in other topics, but I didn't encounter a thread specifically dealing with what I'm looking for.

So the last couple of days I've been thinking about the differences between Ne and Ni. Seeking to understand them in their isolated forms I looked up various descriptions, threads dealing with the differences, visual representations of the functions... and although I think I have a good idea about the 'manifestations' of each function (Ne: 'seeing possibilities', Ni: 'shifting perspectives', 'sudden realizations'), I haven't been able yet to integrate all those tidbits of information into a model that makes sense to me. o_o

The reason for making this thread is that I'm trying to figure out which one I *really* use, as opposed to explaining my behaviour parting from the assumption that I use Fi and Ne. Out of the two, Ni seems to be the most mind-boggling, but that doesn't necessarily mean I don't use it, as it could simply arise from a lack of understanding. Se and Si are both easy to understand conceptually, but obviously I don't use both of them at the same time.

Some threads on this forum and typologycentral.com have been very helpful to me in understanding the differences, but most users who commented in these threads were NTs, so I couldn't *totally* relate. After having read their insights, I'm more interested to hear how XNFPs and XNFJs understand their Ne / Ni, and how these functions concretely manifest themselves in their lives

_______________________

My current understanding (or lack thereof) of the functions:
- Ne: external stimuli triggering certain threads of thought or associations. Unlike Si, the triggered associations aren't usually concrete bits of information that are directly related to the stimulus (for example: someone mentions a Twingo => urban car, manufactured by Renault, price range, people I know who own one, places where I can buy one, in what colours is it available), but ideas or scenarios which are related in one way or the other (someone mentions a Twingo => "if I had a Twingo, I would paint it purple" or "wouldn't it be cool if James Bond drove a Twingo with built-in rocket launchers in his next movie". Or perhaps the word 'Twingo' would remind Ne of the word 'Bingo', which would remind him of games of chance in general, which would remind him of his latest statistics class...).
Ne can serve to combine two seemingly unrelated pieces of information, or can serve to use one property of a bit of info as a trigger to jump to an 'unrelated' bit of info (as in my example: the fact that Bingo rhymes with Twingo permits our Ne user to jump from the realm of automobiles to the realm of odds & probabiliy).

I visualize Ne as a little squirrel, sitting in a tree with almost endless branches. The squirrel can go from the trunk of the tree to the end of a branch, without knowing its destination. Confronted each time with various possible paths, the squirrel will simply choose one on a whim, or decide to sit still and evaluate the different possibilities. The squirrel can also jump from one branch to a totally unrelated branch, as long as they're close enough to eachother.

- Ni: external stimuli are deconstructed into various tiny fragments. These fragments are not put into boxes and 'filed' as Si would do; instead, the 'pieces' are all put together and combined in a whole new way, resulting in an ever-changing, fluid mosaic, this mosaic being the peculiar world view of Ni-users which they draw their data from.
Ni users can choose to direct their focus to 1 piece in the mosaic. The meaning of the piece depends on the surrounding pieces. A blue piece surrounded by other blue pieces is not the same as a blue piece surrounded by red pieces. Because of the fluidity of the mosaic, Ni users can see how the meaning of a piece changes when the surrounding pieces change...

After re-reading my improvised analogy, I feel I've complicated things instead of clarifying them (for myself and others), but yeah... I'm trying to understand this on a conceptual level...

The easiest thing to understand about Ni, is that it enables people to look at things from various angles or perspectives. From what I've gathered, Ni-users consider our perception to be inextricably linked to our biases and assumptions. They try to shake this up by 'redefining the rules of the game' every once in a while, throwing off certain assumptions and looking at things in a different light. Could this, perhaps, be a good description of the Ni way of evaluating a problem... "Playing the same game several times, but using different rules each time"?

_________
Thoughts? Personal experiences? Any concrete examples, analogies or abstract ideas that will help me (and others) differentiate between these two pesky buggers? Post them plz.