Add simple binaries to stellar models

[karllark]

The goal here is to add simple binaries to the stellar models. These would be physical binaries having the same metallicity and age. The parameter space could be/should be reduced further by only including binaries that produce SEDs in the target bands that are detectably different than single stars (e.g., the most luminous of the two).

[mfouesneau]

I think playing with mixture models is the best way to go. This means multiple "grids of kind of models"