In 2013 I have conducted a few more GTL reviews, which bring my total to about 45. Most of the reviewed games are in the 2k-6k range, say 4k on average. Looking back at those reviews, I observe a few returning comments. These are both things a 4k does not do correctly and/or I think I know how to play correctly. Clearly there are many things I don't know and I'll even give wrong advice sometimes.
The most common mistake on a strategic level is that players leave their weak groups and those of the opponent alone to play in open area instead. This leaves or misses an opportunity to attack. Many players overestimate the value of playing somewhere first and underestimate the value of making a group strong or harrassing the enemy's groups. I think this mistake stems from the mistaken concept that Go is about territory in the sense of surrounding open area, while Go is about territory in the sense of covering the board with one's own stones. In the second perspective, existing stones are just as valuable as empty points, while in the former, empty points look more attractive.
Often playing elsewhere goes at the expense of a cut. But even local plays are oblivious of the potential to cut or connect. I am still unsure what is the root of such missed opportunities. Sometimes it seems that a 4k will superficially consider stones to be connected. Probably they do not look very actively for a cut because the value of a cut is underestimated. I think dan players get high (emotional) satisfaction from cutting and connecting so they will more keenly pursue such opportunities.
In close combat, one still observes hane at the head or not extending against a shoulder hit. One sees failing ladders, cutting stones in atari which can be saved but are not, ladders where nets are available, not connecting against a peep while the connection is good ...
I believe the root for this issue is some misguided desire to resist. We do not like to be forced into doing something. We like to play originally and not just apply a lame basic technique. But Go can be played at a very high level if you just apply basic stuff. Many kyu and dan players shoot themselves in the foot by not playing the obvious answer, when it is most often and clearly a good answer.
Sometimes the tactics are as easy as Black-White-Black and then counting liberties. Even such basic sequences are often dismissed when they actually give a good result. I can only think of sloth as the root cause for this. Close combat and especially cross-cuts need to be read out. Apparently the effort is often daunting, while it takes just a few moments of thought.
Some players will include questions in their game and even a few variations. When I was making the leap to dan level, I analyzed my games thoroughly. I played a few slow games with a friend of mine, in a system of shifting handicap. It was very competitive (up with that handicap!) I wanted to really understand what I did wrong and improve on it. I think this kind of self analysis is more important than advice from another player. You need some fresh ideas from somewhere, but that can be a book or any game by strong players just as much as a reviewer.
I got many thanks for my reviews but often I wondered who had gotten most benefit out of it: the reviewee or the reviewer? Many reviews would be more helpful if the reviewee had included an analysis of his own first. It would probably even be sufficient.