test(cover): Intermittent daemon loop cover hole

When running the tests for the different Python versions in parallel, sometimes the
inner daemon loop would take long enough that it exceeded even the `0.1` polling time,
leaving an intermittent coverage hole. I was able to reproduce this on my laptop and I
think on my home server used as a GitLab CI/CD runner.

I'm concerned that this poll value may still not be conservative enough, for example for
someone doing local development on a Chromebook or a tablet. Though such developers are
much more likely to use CI instead of running locally. So we'll see if this intermittent
coverage hole re-emerges and deal with it then.

Author: rpatterson

src/prunerr/runner.py

@@ -344,7 +344,10 @@ def daemon(
             # no event we can subscribe to that reliably determines disk space margin
             # *as* the download clients are downloading:
             if (time_left := poll - (time.time() - start)) > 0:
-                time.sleep(time_left)
+                # Intermittent coverage hole here when the machine running the tests is
+                # slow or overloaded enough that the inner daemon loop takes more time
+                # to run than the poll value used in the tests:
+                time.sleep(time_left)  # pragma: no cover
             logger.debug("Sub-command `daemon` looping after %ss", time.time() - start)
 
     # Methods to list the download items in each life-cycle stage: