I think the criticism was that the number from the metric was given the wrong unit (percentage of pageviews vs. percentage of desktop computers), rather than th

I think the criticism was that the number from the metric was given the wrong unit (percentage of pageviews vs. percentage of desktop computers), rather than that the metric itself is fuzzy.