This is good commentary on value added data. But the explanation for why we get all exercised about this sort of thing is simpler than that. I think what we’re really getting worked up over is the gap between our ability to describe (and prescribe) large scale and small scale behaviours.
Behavioural sciences are very good at describing large scale, group behaviour and notoriously bad at doing the same for individual behaviour. Teaching, alas, is one of those professions that straddle both worlds: on one hand, teachers are part of organizations that work collectively to produce collective results, e.g. broad literacy, a ready work force etc. for society; on the other hand, teachers work to help individuals achieve their best.
We should recognize that the demands of these two objectives sometimes compete and that data is never neutral–it sits in one of those domains or the other. The sheer volume of data organizations such as Google or Amazon collect and their powerful algorithms are getting us closer to bridging this gap, but we’re not quite there yet, certainly not in educution which just doesn’t have its hands on the tools.