Remove charge dependency in NN ghost killer

Hadronic tracking asymmetries are observed in the 2024 data (ref: https://indico.cern.ch/event/1512219/#7-hadronic-tracking-asymmetrie).

From local studies with MC samples (checking the tracking efficiency for charge positive and charge negative long tracks with and without NN ghost killer), there is no evidence that the NN ghost killer is causing the asymmetry. However, as @mveghel recommended, it would be safer to remove the charge-dependent input to the NN ghost killer.

FYI: @mveghel @dovombru @msaur @ascarabo