Bad for the psyche? I've heard that explained as why the functions have to be "ordered" the way they are, and why "dom/tert loops" aren't good for too long, etc. but as Beebe himself says, too much cannot be made out of the order, and as Lenore (or at least the "Exegesis wiki" on her) says, it's not really about functions being "used" like gears. (if it were something like physical gears, then "damage" could occur).
The functions are perspectives, where data is basically divided, between tangible and conceptual input (information gathering), and technical or humane output (decisions). Fi and Se are basically reflections of Ti and Ne, using the same attitudes, and same brain hemisphere, in this case, right, and bearing the P attitude. They actually bear some things in common through the P attitude (hence, why J/P was able to be made a standalone type factor). Normally, Fe and Si will reflect them, but in the opposite attitudes and hemisphere, and the J attitude.
So there is not that much of a conflict between the functions in themselves. It's in the ego structure of complexes where we prefer one, and repress the others, which then seem to conflict. However, there are different complexes, and when these are constellated, they are felt through the perspectives of the associated functions. Hence, any one of them
can come up, regardless of conflicting with the preferred two.
So Lenore's theory in the book (conceived before Beebe's order was widely known) looked at it from the angle of the brain alternates being the first we run to. This still focused on the behavioral aspect of the functions, and is even evidenced in many people's cognitive process (Keys to Cognition) test results, where Beebe's #8 is often third in strength (remember, it is not about strength), and the tertiary and inferior are weakest. Later, she moved away from that, and focused on the Tertiary Defense more.
Here is a diagram that shows both models working together: