Cache/save baseline simulation run

[anth-volk]

Currently, we don't save baseline simulation runs independently of any corresponding reform, meaning that for most users, we're recalculating the same baseline policy impacts. We should determine a way to cache/save these output values independent of the broader data lifecycle discussions we've been having.