Have you considered that thing about facts yet? Still haven't seen any TPs address the facts of facts. It's written into typology that you'll have a weird approach to what's "there" - that jolly Si function, which by definition warps immediate perception, making it inaccessible to the perceiver, and presenting them with something that better conforms to past perception....
And Ti itself - it's built in that immediate external
distinction is unavailable. To distinguish between two items, you need a principle, not a simple observation, right? And this is especially complicated when the items are not immediately physical. If there's some conceptual or modeled item to distinguish from another, do you rely on principles and definitions or on working backwards from unchanged
direct observation that must alter the model before it itself can ever be altered?