I'll just jump right in:
I've been trying to distinguish Ni from Ti. This has been the main focus of my dominant functions, as I am mainly having trouble deciding between IxTP and INxJ. Ti/Ni (dominant?), from my understanding, both break any scenario, situation, or system down; but Ti looks at every piece, categorizes it, and analyzes the parts as a whole, and when combined with something like Ne, Ne generates new possibilities and connections related to the analysis of a part (or could that be Si?) which CAN come out of nowhere sometimes (Like, I could be studying the design and layout [both internal/external] of an orange. I notice that it is soft and tender, so Ne says: "Cut it open!". I cut it open into halves, and Ti discerns what is inside, notices that it is juicy and soft, then Ne says: "Squeeze it so liquid comes out, so you can have juice!". Or is that Ni+Te working together? [Ni looking at every different angle, and since Te is a Je function; orders action to be implemented upon analysis of angles? Hmm.]) From my understanding of Ni, it is similar to Ti, but instead of analyzing everything, it synthesizes everything; combining everything, all connections, visions, external information, from the past and present, and forms a hypothesis upon everything synthesized, with no limit to consistency or inconsistency? I can't really come up with an example of this.
Ne-Si axis: Si is a database, a library of info and Ne calls Si to help whenever it encounters a new situation, to find information possibly related to the situation. Si does it's job, and Ne takes that info and branches out ideas, observations, and possibilities, either in a branch, or interconnected, shifting from idea to idea, all in relation. An example of the former would be this: Cup, drink, soda, fizz, caffeine, rotting teeth (that comes from Si recalling info about how too much caffeine can eventually rot teeth) ugly/unattractive --> which branches out to either "unattractive mouth and teeth" or "ugly face" ---> branches out to "ugly because of teeth" or "ugly in general" ---> comes up with possibilities relating to the external world making judgements of person being ugly, etc. Then we come up with the dots being connected: Greg drank liquid from a cup, which happened to be in the form of soda, which contains fizz, which is actually part of caffeine. Eventually Greg drank so much caffeine that his teeth started to rot, so he chose not to smile often because his teeth would make him look less attractive....blah blah. Or is this Ni combining the observations together? (this is a horrible example, I might come up with a better one later)
Ti-Ne axis: Ti is constantly analyzing every piece of data, from every angle, but Ne looks at them and says "Whoa, that reminds me..." and throws in new abstract data, or says "How about this! Analyze this, and that, or how they relate!" etc. Ti has to then analyze the new data, with no limit to the speed and capacity it discerns (in other words, it can take it's good old time). This can sometimes be an unending process? I'll come up with an example later.
Ni-Te axis: Ni synthesizes info, but can't summarize it into words; it's all a blur of information that can be understood internally. Ni goes to Te to try to clarify it all. How it does this, I can't clearly guess, but I would assume it would be done by Te requesting to look on the internet for words or summarizations that best describe this intuitive blur of understanding? (or would that be Se?), or Te would summarize it as best as possible, and if in discussion, Te would ask a question like "So what you're saying is X is related to Y, which is caused by D which is the result of Xa?" to try and get clarification of their understanding, instead of just point blank asking "Why is X doing this? Elaborate..." which is more Ti? I can't discern this axis very well. I don't know much about Te, though, so that's likely why.
I'll try and come up with more later, but how far off or close to the mark am I with this?