This page lists a few thoughts of mine about teaching Go in dedicated personal lessons. It does not refer to public reviews or writings.
Advice from stronger players is widely assumed to be automatically beneficial. In my view, that's mostly wishful thinking. There are various pitfalls and sources of confusion a teacher should familiarize her/himself with, to allow realizing the pupil's true potentialities.
Supplementary stuff (notes to myself):
If this site is free for comment... sonst magst du es auch löschen. Tapir
You write: it's hard enough for the teacher to guess, or understand, what the pupil wants or what s/he's actually concerned with. So, it can be very useful to a teacher to take care about every hint about it he can get.
Actually, a teacher may ask the pupil before a lesson what he's concerned with. He even should do so... You can then continue teaching by asking. The single most effective method of mine is asking "What did you plan to achieve by this move?"
blubb: Yep, I ask that, too. It often helps both sides.
What I found most helpful was to play a game (which may be left unfinished) with the pupil first, commenting on it later. Seeing someone in action, later demonstrating/discussing lines you would play, or going back and switching colors for that matter may be very helpful. For sure you should not pretend that you are always right... for this is my learning experience, I read the game comment of Toshiro Kageyama at the end of his famous book, and said wow... He comments on a game he won against the Meijin, saying this was his best game ever and he still comments on his mistakes. This gives so much confidence.
blubb: Actually, I don't like the notion of "mistakes" in Go very much. It suggest one could reasonably expect to play perfect moves throughout a game, which, in my view, would be pretty silly an expectation. Regardless of whether you are 30k or 9p, it makes sense to strive for the best you can do - no less, no more. Anyway, I see your point.
But there is a disagreement as well, I disliked your comment on your mainpage somewhat. I don't believe it is necessary to be 10 grades stronger than your pupil. 10 grades may be enough to pretend you are right always, but you still are not, and it may be the wrong teaching method from the beginning. (Intimidating more than encouraging.)
blubb: Oh, I didn't even imagine that dialog could be misunderstood so badly. If you have read it all, together with this page above - sorry, then I perhaps can't help. Not only is every teacher's understanding of Go inherently limited, but it is essential for him/her to be aware of those limits. What I'd call good teaching, takes more attention for the pupil than a teacher usually can afford if the game is so tough s/he's being kept busy with finding the next move alone. To really sense and meet the pupil's needs, to be fully available for questions, you need to actually be free and able to care. Intimidating the pupil with your perceived superiority couldn't be farther off.
Tapir: I have some trouble to connect the dialogue there, to the good advices here. Maybe it is good enough, if the teacher demonstrates what he would do and why he thinks it is better instead of proclaiming this is bad, you shouldn't do this, (the demonstration may improve teacher and pupils positional judgement at te same time)... this is a bad habit, which is somehow common in my experience in teaching joseki etc. And you can do this with people of your strength as well, as with pupils weaker than you... not only with pupils at least 10 ranks weaker. On the contrary quite often a player so much stronger tends to explain too advanced concepts instead of adding to the basics step by step.
And during play, actually there is no need for the teacher to show superiority all the time. Playing e.g. 9x9 with white against somehow only some stones weaker it is not necessary at all for the teacher to win! You can create 2,3 nice tsumego in such a game, comment on it, discuss several moves what they accomplish... i don't mind losing in such a situation. Actually what puzzles me is that you're describing some ideal teacher who cares, encourages etc. but is so far off (10 stones) that the mere appearance of such a superiority may be intimidating as well, you couldn't even see your progress through reduced handicap stones, an occasional win etc. when your teacher is so much stronger. 2nd you may need a teacher after 3k as well... That's in a way what I meant with the Kageyama example, he is so much stronger but still makes you feel that you can be strong a day too, still making bad moves, or mistakes, but ever smaller ones.
blubb: It looks like you and I, we are referring to different degrees of a game being a "teaching" game. Of course, games between players just 2 or 3 stones apart can turn out to entail a lesson or two. In fact, people of similar strength can learn a whole lot by merely playing with each other and rewieving their games. However, in the dedicated kind of lessons I am referring to, the main focus of the teacher is the pupil rather than who wins. The teacher doesn't win all the time there, either, but that's mostly a side effect of the teacher adapting to the pupil's play.
On the 3k pupil: Right, I made that suggestion of a 10 stones optimum when I was KGS 9k, and obviously, the optimal rank difference (see below) gets smaller at higher ranks. I suppose, it would roughly correspond to a particular win ratio if the teacher and the pupil were to play competetively (using their "teaching" settings). A teacher who could win, say, 95 % of those games, may be sufficiently candid and neutral. (Whether s/he really is, that's a different kettle of fish.)
For a particular pupil's advancement alone, the strongest possible teacher is naturally the best choice (ceteris paribus). A strong teacher is free to focus on the pupil, and the pupil can learn straight on for long. From a global viewpoint though, if they are too far apart, that's wasted ressources, because the teacher could likewise teach someone in between, who again could teach the other pupil, without much loss of quality in the transfer of knowledge and abilities.
On the other hand, if teachers are not enough stronger than their respective pupils, quality is affected. Not only is such a teacher too involved in dealing with the actual play (whether or not s/he'll win), so there's not much room for him/her to guide the game and create situations deemed helpful for the pupil, but also, the learning path of the pupil will be meandering unnecessarily.
Of course, these figures are very general; there are likely to be exceptions.