I think the problem here is that such a system seeks to evaluate teachers as if they were line workers, cogs in the machine. In reality, teachers operate more like managers. As anyone who has been involved in management or management education should be able to attest, getting a good read on exactly what makes good managers so good (and bad managers so bad) is a lot harder. The metrics are a lot fuzzier, and there tends to be a lot of different ways to get good (or bad) results. In many cases, two people doing things that look on the outside to be very similar can lead to wildly divergent results.
Go to any random business school and take a look at their various case studies on managers. It's usually quite difficult to find any common thread in any of them, other than "this guy's company was successful, therefore what he did is the right way to do it." Of course, in every one of those studies, the manager did things differently than the other managers. The upshot of it is that the best managers are unique snowflakes who follow their own rules and are successful, while the worst managers are unique snowflakes who follow their own rules and aren't successful
In short, why does Bill Gates think business can help evaluate teachers (leaders of students) when business isn't even very good at evaluating their own managers (leaders of corporations)?
