Particle Flow bug fixes to content of Global Particle Flow variables and FPE
This includes update to the logic that determines the content of two new variables needed for Global Particle Flow (GPF) - the list of matched xAOD::CaloCluster, to each charged PFO/FlowElement, that had charged energy subtracted and the ratio of new/old energy. This MR also includes a fix for FPE that is documented in both ATLASRECTS-6404 and ATR-23684. Extensive additional DEBUG statements have been added to the code for future debugging purposes. Also includes a bug fix to the trigger pflow logic (which is already in master in fact, but was needed here to allow me to run q221/431 on this branch). Testing Done: RunTier0Tests.py shows no problems, except diffs in the output AOD/ESD. Once I exclude the expected differences in the two GPF variables from diff-root no differences are seen. Ran physics validation on 2000 ttbar events and saw no (note the two GPF variables are not currently tested in physics validation) differences in pflow, nor pflow jet, variables compared to default master. Detailed testing on a number of ttbar events shows information for more complex track-cluster systems is now correctly stored.