Use interaction region condition in all tracking and PV finding

A condition with the interaction region parameters was introduced with Detector!383 (merged) LHCb!4173 (merged).

It was picked up in the main PV finding algorithms with !3552 (merged) but multiple uses of DeVP::beamSpot() remain, see below. We probably need to switch to the interaction region condition everywhere.

A few related questions/comments:

• Allen!1261 (merged) did the same for Allen, but does tracking there assume some beam spot position or correctly picks up the new condition?
• TrackBeamLineVertexFinder (the non-SOA version) is not used, not even in tests. Should it be removed or do we need to keep it as a reference implementation? In the latter case we should add tests.

Relevant MRs !3383 (merged) Allen!1085 (merged)

Rec/Tr/PatPV/src/TrackBeamLineVertexFinder.cpp
304:  Gaudi::XYZPoint beamline = deVP.beamSpot();

Rec/Tr/PatPV/src/PVOfflineTool.cpp
338:  Gaudi::XYZPoint beamSpot        = deVP.beamSpot();

Rec/Tr/PrKalmanFilter/src/KalmanFilter.cpp
214:                                            deVP.beamSpot() );
223:                                       is_long_in_beam_pipe, deVP.beamSpot() );
232:                                       *extrap, unique_id_gen, is_long_in_beam_pipe, deVP.beamSpot() );

Rec/Tr/PrKalmanFilter/src/KalmanFilterTool.cpp
619:                           *lhcb.geometry(), extrap, is_long_in_beam_pipe, deVP.beamSpot() );
710:                                            deVP.beamSpot() );
718:                                       extrap, unique_id_gen, is_long_in_beam_pipe, deVP.beamSpot() );
727:                                       extrap, unique_id_gen, is_long_in_beam_pipe, deVP.beamSpot() );

Rec/Tf/FastPV/src/FastPVFinder.cpp
61:    Gaudi::XYZPoint beamSpot = deVP.beamSpot();

Rec/Pr/PrPixel/src/VeloKalman.cpp
76:            fitBackwardWithMomentum( loop_mask, tracksVP, idxVP, qop, hits, 0, devp.beamSpot() );

Rec/Pr/PrPixel/src/VeloClusterTrackingSIMD.cpp
1179:                         ( ( devp.beamSpot().x() - x0 ) * tx + ( devp.beamSpot().y() - y0 ) * ty ) / denom );
1224:        closestToBeam.transportTo( closestToBeam.zBeam( devp.beamSpot().x(), devp.beamSpot().y() ) );
1237:        closestToBeam.transportTo( closestToBeam.zBeam( devp.beamSpot().x(), devp.beamSpot().y() ) );

/cc @mgiza @ahennequ @ausachov @adudziak @wouter @dovombru @mveghel