Have quantum outcomes been tested for the kinds of statistical not-quite-randomness that PRNGs have? Aren't the measurements far too imprecise?