Some have written about the problems of the piece. I agree with most of the criticisms, but what is useful for the rest of us is not the problems, but figuring out what useful things we can get out of the piece.
The author's biggest work-related mistake was to sit tight when given a managed he had a bad relationship with. In most large organizations (500+ employees), your success and your failure will have a lot to do with a good relationship with your manager. It's not just performance reviews, but your technical influence that will be influenced by this. This matters with sibling teams too, as a political manager (and most bad managers are very political) will not want a report they don't like to have more influence than they do. I have seen people go from being seen as extremely productive to being called very low performers, and vice versa, in 2 months and a manager change. This kind of match, and caring about making teams cohesive, is something that most large valley companies do not really care about, and hurts them greatly, as a lot of talent is underused or downright shoved out the door.
We can learn the some things as managers too: A report we really dislike could really help the company a whole lot if they were in an environment that matches them better, and ultimately that's what we should be caring about, not making people we don't like leave, and helping people we like get promoted. Whether it is by knowing them better, or helping them move somewhere else, is far more work than just undermining them, but its ultimately the right thing to do. What we should foster instead is teamwork, and the minimization of political behavior among reports. The sly report that is always telling me what I like to hear, but playing politics all day is the bad manager of tomorrow. Companies work better when people are aligned with building the best product the company can, not maximizing the credit they get. The more energy is spent on political fights, the worse the product gets.