QMM final output should not move the peak markers

[RadixSeven]

This was part of #57

Right now the final output routine does an alignment that moves the peak ppm markers and the identifications. The deconvolutions can be made parallelizable and non-interfering (so you don't end up redoing overlapping bins because a previous deconvolution moved the peaks) if they don't move the markers. This would also make things more predictable for the users.