I wonder if you are trying too hard to find a sharp
distinction? Greek and Sanskrit inherited from PIE two sets of verb
endings, active and middle. The basic distinction was in line with what
has already been posted (middle implies involvement of the agent in some further
way), but in many verbs the use of the middle had become mere habit by the time
of the languages we know. We cannot press either Greek or Sanskrit for a
hard and fast distinction in all cases, even though with some verbs a
distinction is clear.
A good example of bad practice is from a theology
lecture I attended. The speaker claimed that the use of the middle was
significant in the New Testament phrase "No one comes to the Father except by
me". He argued that we come to the Father for our own benift, hence the
use of the middle. In fact, the language has no choice at that
point. The verb exists only in the middle, and by New Testament times the
middle in that verb has no particular significance whatever.
So it is with the earliest Greek and Sanskrit we
have. Some instances of the middle have significance, some do
not.
Peter