Pair Programming is just another tool to add to the set. It works very well with the right environment and the right situation, the same as every other tool/met

Pair Programming is just another tool to add to the set. It works very well with the right environment and the right situation, the same as every other tool/methodology works in its own environment. The trick is getting the combination right for your team/technology/problem. So why does the author feel the need to bring things to extremes? Professing that peer programming only work for the minority and supporting it with anecdotal evidence reeks of confirmation bias to me. There is such thing as a middle ground.