Longterm: support stitching simulations

[eaton-lab]

We talked about this a while back, it seems there are a few really novel applications we might be able to support by running simulations back until some point, then freezing the state and reconfiguring the simulation, and restarting from the saved state:

1. Different recombination rates/maps applied to different spp tree edges.
2. Different mutation rates applied to different spp tree edges.

This is not critical, just long term thinking.