nice article, why don't bosses read it?
Management is a fundamentally messy and error-prone process, and dealing with differences in abilities is always hard and complicated :-/ The other thing about running organisations seems to be the "Anna Karenina principle" - all organisations that function really well are similar, but almost each organisation that fails fails in a different way.
As the paper is addressing a social problem rather than a technical one, the definition of what a 'real solution' may be is debatable.
I definitely think the paper provides suggestions about how to reverse such a cycle.
As a manager, it is normal to be more directive with someone who is underperforming due to (perceived) lack of knowledge. You do need to be clear about you doing so though, and be clear about the path away from it.
I had a manager a long time ago who did it in a nice way. Our office was underperforming compared to the other offices he was managing, so he explained to everyone in an all hands meeting that he would be spending more time on control activities with the goal of getting back to not doing so once we were hitting delivery targets once more. No-one took offense to that, it was more "Ok, we are not performing at the level we need to be, and that means we will spend more time on control activities for a while, but to get out of that and back to our normal more laid back approach we need to do X, Y, Z."
The harmful situation is when you start spending more time on control activities without expressing why you are doing it and without explaining what the path out is. The responsibility for that is for sure on the boss, but also on the employee. If you are not asking "Ok, I have noticed this, why is that?" as an employee as well, then you are contributing to the situation and the downward spiral.
If you have the ability to significantly change someone else's life, they will produce submissive and defensive behavior just as though you had them actually pinned to the ground while baring your teeth. In the end, a manager is just a skilled laborer with a different set of skills.
There is no particular reason why the ability to organize a labor force should automatically result in authority over it. The manager does not always need to be the boss. I imagine that some business decisions would turn out differently if they had to be backed up with evidence and argument rather than authority and the threat of job loss.
Ability to work independently isn't the same thing as good performance either - comparing myself and a co-worker: I work best when given an outline of a 'business' problem and asked to create a technical solution, he works better when given explicit requirements with frequent checks that what he is making will solve the problem. But he writes faster, better code than me, and can quickly solve technical problems I'd find very difficult because he really knows his stuff.
My perception is, a lot of non-technical managers rate ability to 'solve a business problem' higher because it is closer to what they understand and value in themselves.
But yeah, some under performance is to do with relationships, but other under performance is due to people doing work that doesn't suit them, or just not being very good.
The latter half of this is the real big deal. Relaxed control as a reward for doing the "right" thing. Too often, especially in the public sector, the control measures are ever-tightening.
It's much more subtle, and it can work wonders for your career.
I try to get into my manager's head. What are they worried about, what do they not care about, when is it best to present a solution, when is it best to avoid them.
Things as simple as planning on when to meet them is beneficial. E.g. if you want to get their actual feedback on something - schedule it early in the day (earlier in the week is a secondary consideration). If you want them to just agree with you, schedule it later the day, preferably on a Friday.
Want to work on a cool project with a cool technology that will advance your career? Plan for it. Research who the decision makers are, find out how to best influence them, build a polished presentation (could be a slide deck, or something else it depends on your company culture), learn the best way to "close" (http://en.wikipedia.org/wiki/Closing_%28sales%29).
I could go on and on, but I hope that this helps.
It's up to you to manage your career. Nobody else is going to do it for you.
Setting expectations is what you do for things that are asked of you. Of all of the things that I have done and gotten "credit" for that have had an impact on my career, very few were things that I was asked to do. Setting expectations is about making sure the orders you are given are reasonable. I treat it as a tool I can use to ensure I have time to work on stuff that is actually important and will have a long lasting impact(not just to me, but also the company, I'm not a complete sociopath).