[WIP] Use the full 64 bits of the cellID for surface lookup

[tmadlener]

Switching to e.g. system:8 bits can lead to cases where CellID0 truncates bits off of the cellID, making the lookup fail.

BEGINRELEASENOTES

• Use the full 64 bits of the cellID to do the lookup for Surfaces for digitization in DDPlanarDigiProcessor

ENDRELEASENOTES

Triggered by the discussion / conclusion in key4hep/k4geo#568 (comment). Without this a simple change to system:8 (from e.g. system:5 will make the surface lookup fail as CellID0 truncates to only 32 bits)

I have tested this with CLD_02_v08 replacing system:5 with system:8 and using the --trackingOnly option for CLDReconstruction.py and tracking runs in both cases. However, efficiency seems to be down by a factor 2 for system:8 at the moment