update g_lineplot snaps (#1157)

fix errors reported on ocean integration tests (both valid and nonvalid)

```
✖ | 2       1 | g_lineplot [2.6s]                                               
  ────────────────────────────────────────────────────────────────────────────────
  Failure ('test-g_lineplot.R:27:3'): g_lineplot works with custom settings and statistics table
  Snapshot of `testcase` to 'g_lineplot/g-lineplot-w-stats.svg' has changed
  * Download and unzip run artifact
  * Copy 'tests/testthat/_snaps/g_lineplot/g-lineplot-w-stats.new.svg' to local test directory
  * Run `testthat::snapshot_review('g_lineplot/')` to review changes
  Backtrace:
   1. vdiffr::expect_doppelganger(title = "g_lineplot_w_stats", fig = g_lineplot_w_stats)
        at test-g_lineplot.R:27:2
   3. testthat::expect_snapshot_file(...)
  
  Failure ('test-g_lineplot.R:43:3'): g_lineplot works with cohort_id specified
  Snapshot of `testcase` to 'g_lineplot/g-lineplot-cohorts.svg' has changed
  * Download and unzip run artifact
  * Copy 'tests/testthat/_snaps/g_lineplot/g-lineplot-cohorts.new.svg' to local test directory
  * Run `testthat::snapshot_review('g_lineplot/')` to review changes
  Backtrace:
   1. vdiffr::expect_doppelganger(title = "g_lineplot_cohorts", fig = g_lineplot_cohorts)
        at test-g_lineplot.R:43:2
   3. testthat::expect_snapshot_file(...)
```

I just re-run the tests and accepted updates for erroneous test cases.
Those diffs are actually good. Interestingly I got more diff than just
g-lineplot (that is: `test-g_km.R` and `test-decorate_grob.R` but those
are probably due to my local setup as we don't have any other failures
reported for those files so I haven't included them