The number of MC sntp files resulted from MC processing do feature a number of distorted shapes, like in "track momentum from range" variable. This issue has appeared first time with MC (daikon_10) files of L100200N configuration produced at the end of Dec, 2012 and notably present in the fluka11 MC sntp files (daikon_07) produced in Feb, 2013 for a number of different configurations. This feature seems does not depend on the flux type (fluka08 or fluka11), neither from the dogwood5 or dogwood7 reconstruction package. Reroot files from MC generation have been produced at TACC. Currently, W&M farm makes a test run with fluka11 flux files and L010185N configuration to check whether the problem has something to do with the MC generation at TACC (in case the MC data produced at W&M and reconstructed at Fermilab won't have the same features as TACC sets).
The diff between two log files from the event reconstruction (dogwood5) is attached here: