I don't know all the inner workings of the Foz, but based on the stated method of crosstalk, I can tell you it is not optimizing azimuth, which means it is not optimizing what the needle can extract from the record.
I guess you should contact Jim Fosgate and help him create a better and more accurate device.
I have done a crosstalk measurement using a DMM after setting with Foz, my results gave me a crosstalk of more than 32dB using 1KHz tone, which lines up with the Lyra spec. Also using the Wally chart the channel difference is less than 1dB.
I'm sure this is not always spot on every time, but it remains the range I get when I do test.
Again, everyone should align with what they are comfortable with and then listen and reset as needed till you hear improvement. Then sit back and enjoy...why beat your head against the wall??
It will never be 100% perfect, analog records and turntables are a mechanical moving physical media, get as close as you can.