[Halcro]

The thing about 'stylus drag' is that it not 'constant'.
As Neo says....stylus drag is due to friction of the stylus traversing the groove modulations and the groove modulations vary with frequency and amplitude. In other words, you will have more 'drag' with very loud low-frequency modulation than with quiet high-frequency information.
The Platterspeed App uses a steady-state 3150Hz frequency tone so adjusting your turntable speed to this will not prove that your turntable does not still suffer stylus drag....especially if the speed of the table is different without a cartridge playing.
The only accurate way to visualise the effects of 'stylus drag' is to play the entire side of a 'real' record with demanding music whilst monitoring the Timeline laser on a wall mark.
Most turntables are simply unable to pass this test with the flashing laser either moving slowly to the left or right of the mark with each and every revolution.

[Wayner]

Tables with heavy platters (VPI, Empire) will plow thru any modulation variant, and there is nothing that can be done about it anyway.

[Halcro]

Tables with heavy platters (VPI, Empire) will plow thru any modulation variant, and there is nothing that can be done about it anyway.

Not true....and proved via the Timeline.

[neobop]

Tables with low torque to platter mass ratio tend to be the worst offenders.  Those are the ones that some of us characterize as unlistenable. 

A heavy platter with good torque and servo control like the Goldmund Reference seems immune to this.  There are other factors like bearing friction, but they all impact on torque/weight delivery.
neo

[neobop]

It seems to me that the Feikert 7" speed test takes into account the effect of the needle on the speed on the LP. So if you're concerned about that, then use this method to adjust the speed!

You would think so, but the Feikert app isn't very accurate.  It's designed to measure wow and flutter, but the 7" record is apparently of questionable parentage.  There are a number of examples of this app on Halcro's "DD are we living dangerously" thread on Audiogon, and also in Stereophile.  In every example the table has a different mean freq. even quartz locked DD models.  These tables are more accurate than the app. 

There is a DIN record which can be centered and it has a locked groove.  This is used with a program and can measure W/F to 0.6% unweighted.

If you want to check absolute speed, I think a laser RPM detector is a better idea, and much less expensive.  It will give a continuous readout to a couple of decimal places with the needle in or out of groove.  They're designed for high RPM like car engines.  Consequently, they're very accurate at low RPM.
neo

[mav52]

So using any of these speed or Wow and flutter measurement devices/apps, what does one do if they observe a speed fluctuation , start changing motors , belts, platters. Its nice to look at measurements but taking it to the next step to correct a speed error can get complicated.