Fix dr and dz for geo sim plotting

[jbeirer]

This pull fixes the plotting of the cells in the simulation tests. The way the dr() and dz() of the cells is handled currently in the geometry is misleading (and wrong) as:

• The width of barrel layers (0 - 3) dr is missing a factor 2
• The width of all layers dz is missing a factor 2

This is the result of an inconsistent treatment of cell widths between detector regions in ATLAS. The values are not used in the simulation and don't affect it. However, it does change the display of the simulated calo showers. For instance:

The missing factor is now temporarily only added in the serialization of the simulation states. In the geometry ROOT file that will replace the current approach, the cell sizes are treated correctly (also see https://github.com/jbeirer/pygeosimplify/pull/135), such that the factors will need to be removed again once we make the switch.

[codecov[bot]]

Codecov Report

All modified and coverable lines are covered by tests :white_check_mark:

Project coverage is 37.92%. Comparing base (496412e) to head (8349a20). Report is 14 commits behind head on main.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #34 +/- ## ========================================== + Coverage 37.91% 37.92% +0.01% ========================================== Files 225 225 Lines 10815 10817 +2 Branches 1551 1551 ========================================== + Hits 4100 4102 +2 Misses 6715 6715 ```

:umbrella: View full report in Codecov by Sentry.
:loudspeaker: Have feedback on the report? Share it here.