evalcast plot_coverage type one needs formatting cleanup

[sarah-colq]

type one plots for coverage plots do weird formatting. They will have the forecaster as the x variable (for example), but then also facet by the forecaster, so we end up with repeated information. If you try to make sure the x variable (grps[1]) is not faceted, then it is still treated as a color variable, so you end up with different colors for the x variables.

Formatting for these plots needs to be cleaned up so that variables are not repeated across two types of aes (facets, colors, and x/y axis), and tests need to be written for the one plots.

[sarah-colq]

Additionally, if as a result of this grp_processing_for_facets is no longer used, that function can be deleted, as this is the only place it is used.

[sarah-colq]

To be clear, PR does not fix issue, just referencing a PR that covers the all type.