This has been discussed numerous times here, and I agree with the premise of this article, but some of the prescriptions for solving the problem just seem silly

This has been discussed numerous times here, and I agree with the premise of this article, but some of the prescriptions for solving the problem just seem silly and reek of over-complicating and over-managing the process. Trying to gauge or limit how much time a developer spends on a given solution is problematic because a) you're professing to know the solution before it has been developed, in which case you should have just written it yourself or told the developer the solution, and b) you're just making it worse on the developer by imposing an arbitrary deadline that will only induce stress and result in the solution coming later, not sooner. As a deadline gets closer, the productivity of a developer is proportional to the amount of time remaining. The only exception is if the solution, or the design of the solution, is known and all that is left to do is the typing.