One thing to keep in mind, is that the graph shows exceptionally bad luck over a very short term. To get a true picture of how a payout method effects a user (and a pool) one must consider all possible outcomes.

When we were developing ESMPPS we created a simulator to plot the various possible outcomes of a payout system. It doesn't have a lot of these methods, because some we were not considering. But, it does currently simulate ESMPPS, PPPS, SMPPS and Prop (for a baseline). And, since it's open source, I highly encourage everyone to play with it and (perhaps) implement some of the above methods that are missing.

The URL is: https://github.com/lowentropy/pool_sim