Fix Z scaling for Freedman et al. opacity tables

[aarondotter]

Freedman opacities assume the Lodders 2003 solar composition. The tables are at [M/H]=-0.3, 0, +0.3, +0.7, +1.0, +1.5, and +1.7. kap/preprocessor scales these w.r.t. Z=0.02. Lodders 2003 Zsol = 0.0133 != 0.02 so the scaling done in kap/preprocessor needs to be fixed.

Unfortunately, they do not say in the paper how exactly they scale He with "M" and/or H so we can't be sure what Z values these [M/H] correspond to.

[jschwab]

We're going to need to run brown dwarf models (which use these tables) as part of testing CMS so it'd be useful to get this sorted out soon-ish.