I was actually looking for where you and FDG stated this, before I made the post. I'm confused about the two kinds of "supervision" involved. Like what does "supervision" mean for each case? One thing that seems right is Smilingeyes stating that the Creating function "pushes" the Accepting version around.
they're both ways of interpreting the empirical phenomenon of supervision between types. what supervision is is not dependent on the interpretation.

however, how i would transcribe the interpretations:
1st interpretation: the "bull in a china store" effect; one person behaves indiscriminately (~Accepting) and without intending it pointedly, ends up harming the idealistic (~Creating) activity of a person using the opposite function; the conflict between the functions is settled by the fact that one person uses the energy-efficient indiscrimate version whereas the other person uses an energy-intensive discriminate version
2nd interpretation: the "criticism and correction" effect; one person uses an indiscriminate approach (~Accepting) to find an answer to a question and is corrected by a person using the same mental tool in a discriminate (~Creating) way

the important thing to realize is that BOTH are going on between any supervisor and it's supervisee.

i like the introduction of the 2nd interpretation because it dispells the notion that Accepting functions are "better" than Creating functions because they are incapable of being attacked through supervision. knowing both interpretations means you have to conclude than ANY function is vulnerable. there are no mental "safe havens", no hiding places where you can get away with "not thinking". only by using the full range of functions is perfection reached. i like this view.