Add Particle PID Scores and Availability of Corresponding Energy Estimation Methods

[alexbooth92]

In the current version of CAFMaker, for a given reconstructed particle, you are given a single PID hypothesis and no corresponding measure of confidence.

Instead, for each reconstructed particle, we'd want a set of PID scores and set of energy estimates corresponding based on different hypotheses.

[chenel]

If we manage to implement DUNE/duneanaobj#23, then it would be possible (desirable?) to work this the other way: we simply store one SRRecoParticle object corresponding to each particle hypothesis. Then, we could write a convenience function which allows the user to get a list of all SRRecoParticles corresponding to a particular SRTrack (or whatever).