I believe I'm talking to someone who has no way what he's talking (or does he?), and wanted to see what others thought of it:

I agree that MBTI does a poor job explaining itself. The only concepts from MBTI that I even use are the four-letter type abbreviations and the functional orders--which are NOT the only functional orders that exist in practice, but rather represent ideal personality balances which provide the most utility.

To me, the functions do not represent individual specific actions, but rather broader value systems which govern the most basic motivations for everything we believe. I am in the minority on this forum because I do not believe that shadow functions are really used at all, but only appear to be used when a similar function reaches a similar conclusion.

I believe that all four functions combine to produce one fluid value system with varying levels of influence from each function. Again it's crucial to recognize that the functions are not single actions but rather complete value systems based on the way in which the world is perceived and judged. In this context, a person cannot use Fi and Ti because these value system contradict each other regarding how internal judgments should be made. Whenever a person uses Ji for anything, if you pry enough you can always discover whether Fi values or Ti values were the root cause--regardless of whether they happen to agree on the surface.

Here is an example of my explanation of shadow functions:

I may "use Fi" sometimes, but not because I place any fundamental value in Fi itself, but rather because I recognize situations where Fi's values happen to align with my own (which are invariably the result of Ne+Ti+Fe+Si.) I have no shame in admitting that I find Ti a totally superior system for internal judgments, but then--of course I do, I'm a Ti user! Again you need to direct your focus toward the total reasoning process and its most basic underlying values, not just the surface behavior or end conclusion.
Uh yeah but you've ignored and completely glossed over my point about shadow functions by simply declaring, "Of course we use them!"

It's my contention that we don't use them because they heavily conflict with the value systems of our preferred functions. An Fi user's entire conception of ethics comes from the inside; the idea that one's ethics should bend according to external standards is antithetical to Fi's entire worldview. This is why I do not believe that Fi users ever use Fe (even when they do things that appear to promote Fe goals--they are only doing them because they serve more important and often unseen Fi goals, not because they see intrinsic value in Fe. That's the key concept here.)

When you see the functions as overarching value systems, it seems ridiculous to claim that people switch between glaringly contradictory value systems on a regular basis.

An Fi user might perform actions that appear Fe-motivated, but in reality you can explain the motivations for these actions using Fi and the person's other three natural functions. The explanation will always end up being reduced to "My internal sense of ethics said it was the right thing to do", not "My external surroundings said it was ethical so I went with it."

Fi users are, on principle, against bending ethics to the external environment because ethics are considered private and personal from the Fi perspective. It doesn't make sense that people would just randomly switch between these two directly conflicting values.

When an Fi user makes an external judgment, it comes out in the form of Te because Thinking is the process that naturally makes sense to that person for external judgments.