As I understand it, its already been mathematically proven that the efficiency of a utility function declines with the search space, and there can be no global

As I understand it, its already been mathematically proven that the efficiency of a utility function declines with the search space, and there can be no global optimum when the search space is infinite. When there is no global optimum it becomes more and more important to have a novelty function that can at least maximize the number of future possible local equilibriums, from the current equilibrium. The search space of "Every possible life form over 4.5 billion years" is very large, so you would expect there to be a large reward for the right novelty function. And this appears to offer proof of such a mechanism.