Make chains stop at a hard limit using the stat_* argument

[jamesmbaazam]

Would an alternative be to ensure it doesn't, i.e. select cases in the final generation such that the stat is stat_max?

I think this might be the desirable behaviour. We could treat the last generation of offspring as potential offspring and sample the number needed to reach stat_max by finding the difference between the cumulative number before the last generation and stat_max.

Originally posted by @jamesmbaazam in #237 (comment)