As I've said before, defense of agile and Scrum can easily get into "no true Scotsman" territory, with tedious back-and-forths about whether you are really following the philosophy and if your processes were just more pure you would see the light. Fortunately, this essay neatly cuts off that debate with "I have literally never seen Planning Poker performed in a way which fails to undermine this goal." This is an argument that the ideals of Scrum are hard to achieve, or that managing people to get them to buy into a new process is difficult; it says nothing about the success (or lack thereof) that would result if you were able to follow the methodology.
For anyone who wants to try that strategy, good luck. I think you'll find that if you actually do scrum (rather than all the aberrations the author describes), you'll find you can do pretty well even without a team of nothing but rock stars.
The one useful observation it makes is that you don't really need F2F meetings, something that has not escaped the notice of the thousands of scrum teams around the world that include remote members. If anyone wasn't already aware of it, at least they got something. The rest pretty much sums up to the comment I made above.
