Accept variant as a list

[scarlehoff]

The reason why this is going to be necessary is that for the pheno paper many grids are being recomputed and they are not necessarily in the same structure as before (e.g., they might be normalized differently, or even be a different number of grids -like for jets-).

Therefore it will be necessary to be able to say variant: ["legacy_uncertainties", "legacy_theory"] so that both the uncertainties and the theory are updated.

That said, if people already (even if the theories are different and the yamldb repeated in many places) would use for the new grids (when they are replacing a new one) the same names that were used in Theory 708 / Theory 700 that will make it more likely that they work.

(and remember, we are separating Z0, WP, WM! There are still grids here and there with all three of them concatenated but they should be considered deprecated)

@achiefa @niclaurenti @andreab1997 @cschwan @t7phy @comane