Good map, but it's missing an axis: how granular the simulated unit is. Everyone here builds populations or segments. We've been building toward individual-level decision actors instead, a narrower, harder problem than population accuracy. Building this at Faultline.
