Include RetinaFullCluster in 2023-hlt2pp sequences
As pointed out in Rec#513 for a correct MC-truth matching both make_VeloClusterTrackingSIMD and make_velo_full_clusters must be bound to use RetinaClusters.
As required (Rec#513 (comment 7708883)), I implemented the simplification of !2786 (merged)
I also remove the binds to get_UpgradeGhostId_tool_no_UT and get_global_measurement_provider since they are not required since by default they use the cluster produced by make_VeloClusterTrackingSIMD.