Why raise the point in the article if the measurements are less than tolerance level?Several reasons come to mind:
1)As I and Kijanki indicated, whether or not a given tolerance for a given parameter is sonically significant will depend on the design of the specific component in which it is used.
2)For a specific component design, a specific part, and a specific parameter, the line separating tolerances that might be sonically significant from those that are overkill will often be a gray area, about which opinions can differ. Obviously the designer of the particular component is in the best position to make those judgments (which is not to say that he or she will necessarily make them correctly).
3)Perhaps most significantly, my perception has been that it is EXTREMELY common for audio-related white papers and marketing literature to cite and describe technical differences without addressing in a meaningful way whether or not those differences are QUANTITATIVELY significant.
Regards,
-- Al