This paper gives a method for *dynamically* probing the split peas in a pod hypothesis. We find a broad distribution of Q values centered on ~10^6. We expect Earth/Neptune planets to be 10^3-10^4.

The current number of planets is not high enough to determine any bimodality, but we find it will only take ~800 more planets with the necessary parameters to tease out any such bimodality if it exists.