RICH Optimisations for a 'Ultimate speed' tuning

Some more speed optimisation driven by a study into how fast I can push the RICH HLT, making whatever compromises required, as a possible HLT1 candidate..

This allowed me to find a few additional technical optimisations, that have no impact on PID performance, but also are slightly beneficial for HLT2. Saves about 3% of the RICH timing, so roughly 20% * 3% = ~0.5% in HLT2.