There was no noticeable positive impact from having coverage reports on
every single test run.
While adding a separate command to check coverage might be added in the
future, the overhead of collecting the coverage for every test run made
seems not worth it currently.