EFTracking small fixes to allow Hough Transform to run successfully

A bunch of small changes to allow the Hough Transform algorithms to actually run on real input wrappers. We had not yet tested this. The full chain still doesn't work super well because we need single muon samples in small regions to optimize the geometry, but even before this the current samples would not run (they had never been tested!) Adding @fpastore, @jmasik and @tbold