The problem is, as often in economics, defining the terms and building a model that approximates life. It's not as trivial as a sorting algorithm or a simple 1 parameter regression.

How do you define wealth concentration, for example? If your definition is flawed, you can throw out the entire model.