And because this sort of big business often focuses obsessively on what can be measured, ignoring what cannot be. Even if the later might be more important.

You can measure the number of vertices in a model, the total resolution, the expected gameplay length, the number of dev hours that went into a project. But you cannot reasonably measure the fun value of your game; at most you can rank it in comparison with other games. So fun value takes a backseat, even if it’s bread and butter.