The B-fields should be placed after the energy fields but before the species fields.
Seems like a big change -- does this make them line up with the field numbers in the hierarchy file?
Yes, all of the field names in the .hierarchy and HDF5 files match the data after I made this change. Before, the fields and field data were mismatched in both files, causing me to be completely confused when initially inspecting the data!
Also in the Grid_MHD3DTestInitialize.C routine, data fields are initialized in this order.