Final changes to fully support dd4hep RICH PMTs/Panels

Requires Detector!158 (merged) Rec!2794 (merged) Panoptes!146 (merged)

@gcorti @seaso @bcouturi @bmalecki @tlin FYI

This MR is now ready to go. With it I have removed all use of detdesc objects when the stack is built for dd4hep, apart from a small set of calls that require volume intersections

• DetDesc intersectionTicks method to get the entry/exit points with the volume associated to an object
• DetDesc isInside method to check if a given point is inside a given volume.

As the reconstruction now fully uses the dd4hep geometry I am running into an issue due to the fact I only have MC samples generated with DetDesc, and as this geometry does not match the dd4hep one, the PID performance is highly degraded. Essentially, I cannot fully validate the reconstruction w.r.t. final PID performance without MC samples generated with a geometry that matches the dd4hep one. @seaso @gcorti any timeline on when samples matching the dd4hep geometry might become available ?

Note, due to some small machine-precision level changes, ref. log updates are expected.