Add focused test coverage for utils_plot helper functions

[Copilot]

This PR adds direct testthat coverage for the plotting utility helpers in R/utils_plot.R, which previously lacked dedicated tests. It targets both plot-construction behavior and utility/data-shaping edge cases for the listed helper functions.

• Coverage added for all requested utilities

  • New test file: tests/testthat/test-utils_plot.R
  • Adds assertions for:
    • plot_timeseries(), plot_error(), plot_aa()
    • average_age_line(), cohort_line(), reference_line()
    • axis_breaks(), cap_first_letter(), filter_data()
    • get_id(), calculate_reference_point(), label_magnitude(), check_grouping()

• Plot behavior checks

  • Verifies returned object class (gg) for plotting helpers.
  • Verifies conditional layer behavior (e.g., horizontal line present/absent in plot_error() based on hline).
  • Verifies expected faceting/legend behavior in representative cases.

• Data-path and scalar helper checks

  • Exercises filter_data() with both data-frame and multi-model list inputs.
  • Validates reference-point extraction (including lbs conversion and missing-reference behavior).
  • Validates string/ID/magnitude/group-detection helpers across nominal and edge inputs.

Example test pattern introduced:

error_plot_hline <- plot_error(dat, hline = TRUE)
error_plot_no_hline <- plot_error(dat, hline = FALSE)

has_hline <- function(p) {
  any(vapply(p$layers, function(layer) inherits(layer$geom, "GeomHline"), logical(1)))
}

expect_true(has_hline(error_plot_hline))
expect_false(has_hline(error_plot_no_hline))