Aside from the different values in BoundaryDimensionType.X, the two files look completely identical to me (same datasets, attributes, types, sizes,…). However, I don't understand why the new file is about twice as big as the original one:

This is probably something simple, but I could not figure it out. 

Any suggestion would be much appreciated! Script is attached and the hdf files can be found there:

Thanks a lot,

JC