More plots comparing intpur and Mes for the new and old production can be
found at:

OldCockIntpur0.6_0.75.eps and
NewCockIntpur0.6_0.75.eps are mes distributions for Intpur 0.6-0.75, with
both the old and new Cocktail samples.
The plots are not normalized to the same number of events as we do not
know the number of generated events for the old production, but the shapes
look the same (and ok).

DataUnscaledIntpur.eps shows the integrated purity for run1+2 data.
The yield is greater with the new production (red) than the old production

This is also reflected in
OldDataIntpur0.25.eps and
NewDataIntpur0.25.eps which show fitted mes plots for data (intpur>0.25)
with the old and new production. The higher yield with the new production
does not give much increase in signal but an increase in background.