I know that Brad,
According to the tests in the old Hifi Exklusiv magazines, you can determine the speed to within about 0.04% using a perfect test record. After that,begins phantasy.
Yes. theoretically the precision of the speed measurement is limited by the frequency of the signal and the time length of the test track when using counters. However recording the signal digitally and using fitting or other techniques based in software we can overcome this limit using phase information.
But yes, IMO going over .1% is a phantasy, unless we operate a radio station and advertisers pay us fortune for each second of adverts ...
The best wow&flutter meters from the past, EMT, Ortofon, had the smallest maximum size of 0.02%. I wonder how they determined this value of 0.003 %
II don't know which edition that was month/year
The old Leader LFM-39A , that was considered the Rolls-Royce of the audio meters, had a wow and flutter full scale range of 0.03% - I think it could easily carry such measurement. I saw these meters going for high values at eBay, I resigned to get the cheap Kenwood.
The manual is available at http://bee.mif.pg.gda.pl/ciasteczkowypotwor/Leader/LEADER LFM-39A Instruction Manual.pdf