Improve PI0DALITZ model to dynamically work out maximum probability to make it usuable also for eta --> llgamma decays. Model ETA2MUMUGAMMA is pure one-to-one copy of the PI0DALITZ and as such it is removed.
In following pdf I compare old code and new version for pi0 --> e+e-gamma, eta --> e+e-gamma and eta -- > mu+mu-gamma. In comparison, old code uses artificially increased maximum probability. Nothing changes on the q2 distribution to the point that the two distributions are hard to distinguish.