plot_short_v_long: Plot bias, precision, and accuracy of metrics
In WDNR-Water-Use/CSLSlevels: CSLS Lake Level Analysis

Creates a facet plot with lakes faceted horizontally and error estimates faceted vertically.

plot_short_v_long(
  df_short,
  df_long,
  df_early,
  metric_name,
  variable_breaks = "sort",
  variable_labels,
  value_title = "",
  convert_units = "",
  point_size = 0.5,
  text_size = 10,
  rotate_xticks = FALSE
)

`df_short`	data frame with the median/sd for hydrologic metrics w/a short time series
`df_long`	data frame with the median/sd for hydrologic metrics w/a long time series
`df_early`	data frame with the median/sd for hydrologic metrics w/a short but early time series
`metric_name`	name of metric to display on this plot
`variable_breaks`	vector with values to use for variable breaks in color legend. Defaults to "sort" to convert unique options in "variable" column to numeric values and sort in ascending order.
`variable_labels`	vector with strings to use for variable labels in color legend. Defaults to "breaks" to indicate use same values as breaks. Can also be "percent" to add a percent symbol to values in variable_breaks.
`value_title`	label for y-axis values
`convert_units`	logical defulats to "". If "metersTOft", converts values from meters to feet.
`point_size`	size of points, defaults to 0.5,
`text_size`	size of text on plots. Defaults to 12.
`rotate_xticks`	defaults to FALSE. If TRUE, rotates x tick lables 90 degrees