Create an updated Free-free model

[zonca]

@brandonshensley @seclark @giuspugl the f1 model we inherited from PySM 2 has sharp features in the template, and causes complications with spherical harmonics transforms.

For example, if we evaluate the model at 20GHz, then we want to upscale it to 2048 using spherical harmonics instead of ud_grade, we get significant ringing.

See notebook here: https://gist.github.com/zonca/c44d70c72b211ca6b9a1b999f68b36c0

The workaround for now is to use ud_grade, which creates some strange features in the spectra at high ell but it is otherwise well-behaved. I don't think it is worth investing time trying to solve this issue.

I think we should prioritize creating an updated Free-free model, even if there are no new measurements, we could use the d10/s5-style high ell extension. Maybe someone in the Panexp group could volunteer to do this, I can help reviewing the code.

[seclark]

Sounds like a good plan to me! Let's bring this up at the next PanEx PySM chat. Maybe you can show this. With the paper winding down I bet we'll have some free energy available.