normalize the datamodel differences for binid and binid_binned_spectra

[havok2063]

We should normalize the differences in the datamodels so that Marvin can handle them without changing the Marvin code people use in their scripts. For example, similar to what we did with template_kin and template for MPL-4 and MPL-5, we should do the same for binid and binid_binned_spectra for MPL-4, 5 and 6. If someone tells me where the appropriate place is to make this change, I can do it.