By definition (and I share the original post's issues above), a subject in this state is already far enough out of calibration that they cannot self-diagnose issues, let alone work towards correcting them. Also, maybe there aren't actual problems with the subject themselves, but rather with the environment in which they are trying to operate.

As with so many things, any defects must be measured from a wider scope within which comparisons can be made. Corrections can also be incredibly difficult or just costly, if a positive direction can even be identified.