Investigate age/phase prior criteria for the fit_masses stage

[SteveOv]

Currently have a "straw man" Gaussian criterion applying a prior of the calculated M-S midpoint log(age) +/- 0.3 to the model log(age).