the phone app and shaknspin are quite consistent on peak wow and test record also( Test record TRS1007 ang W&F software WFGUI)
Everything is relative I guess so if I can get the variations down and then use HR supply to set correct speed I am happy-
I use a light puck to put the phone on to have it over center of the spindle, so the phone in in center, Shaknspin also
I am waiting for the Iphone app to plot curves form Shaknspin. I believe Shaknspin should be the most accurate method, removing testrecords and eccentricity form the equation. The most confusing is the terms and definitions. The phone app shown the peak deviation and call it wow&flutter which it is not. Shakenspin calculate many parameters.The best thing is actually to look at the curves
I also started with 0.2% peak Wow& flutter, but spending a lot of time experimenting brought it down. Also without a good belt low values are not possible, I use a 15 year old belt, all new belts give worse results. A non-Michell Nitrile belt gave me 0.8%!!
Another days data , cannot be compared with above. Test record with eccentricity. demodulated woe&flutter in Audacity, not sure how to interpret scale