Fix inconsistent usage of particle mass definitions for secondary particles coming from PROPOSAL

Within proposal::InteractionModel::doInteraction(), we loop over all secondary particles produced by PROPOSAL (here) and put them on the stack (here). Note that the energy returned by PROPOSAL is a total energy. To calculate this total energy, PROPOSAL used its internal definition of the particle mass. When putting the particle on the stack, CORSIKA 8 transforms the total energy back to kinetic energies, but uses the particle mass definition of CORSIKA 8. This is inconsistent.

On the test-high-energy-showers branch (commit f4e794b4), when switching maxRadians to 0.1, using proposal/7.5.1, setting the emcut to 0.5 MeV, running radio_em_shower.cpp with 1e4 2723141261 crashed with an arithmetic exception. The reason for this was that PROPOSAL simulated a compton energy loss, where a photon transferred a very small amount of energy (~eV) to an electron (note that this is very rare). The electron mass within CORSIKA 8 is defined as 0.511MeV, within PROPOSAL as 0.5109989461MeV. Due to the inconsistent usage of the mass definitions, the electron on the stack ended up with a kinetic energy smaller than 0.511MeV (but larger than 0.5109989461MeV). This caused the crash within EnergyMomentumOperations::calculate_momentum.

With this MR, this inconsistency is solved.

corsika/detail/modules/ObservationPlane.inl

@@ -51,7 +51,7 @@ namespace corsika {
     Vector const displacement = pointOfIntersection - plane_.getCenter();
 
     // add our particles to the output file stream
-    double const weight = 1.; // step.getParticlePre().getWeight();
+    double const weight = step.getParticlePre().getWeight();
     this->write(step.getParticlePre().getPID(), kineticEnergy, displacement.dot(xAxis_),
                 displacement.dot(yAxis_), 0_m, step.getTimePost(), weight);