Improved Ring Reuse in RICH Mass hypothesis ring creator

Enhancement on the changes made in !2007 (merged)

With this MR when the same CK ring is to be re-used by more than one hypothesis, instead of just using the CK theta value for the lightest ring, an average is computed and used instead. Offers a slight improvement on the PID performance (essentially no drop from the no-reuse scenario) with at the same time allowing for a slightly higher % of reuse. Reuse tolerances have been tuned to 5%/10% of saturated CK theta values for RICH1/RICH2 respectively.

I also in this MR add a new monitor for the mass hypo. rings, and tune a bit the histograms for the track material monitor to be a better fit for the upgrade detector.

Brunel ref. updates are in Brunel!1011 (merged), and Moore ref. updates in Moore!485 (merged).

Please do NOT squash the commits here, as !2058 (merged) is based off of this branch.

Requires LHCb!2521 (merged)