"Productivity" is the wrong metric to measure for developers anyway. Development is not a factory job, developers aren't cranking out widgets on an assembly line and developers who crank out more widgets aren't somehow superior to those who crank out fewer.

What matters is the value of what devs have created, and that's almost impossible to measure objectively in many cases, with some occasional exceptions. That often takes a lot of work and usually requires someone who is also highly skilled making a subjective judgment. I complete agree with the author here, hiring and culture is hard, and most of the time that fact is ignored and the result is failures on both counts. Taking the time and effort to be good at either is usually a huge competitive advantage.

It's similar to software. Does your product solve easy problems or hard problems? If you only solve easy problems, how do you expect that to keep you ahead of the competition? Determining the talent of a developer is not an easy problem in the general case, and if you only hire the developers that are easy to hire you're probably doing it wrong.