I don't see the difficulty. The mechanism for PA on the EQ6R is exactly the same as machinists/engineers use to adjust precision machines, since forever. If you work with principles of the mechanism and mitigate the limitations of manufacture ie: shim the AZ rotation boss with shirt box lid type plastic to remove slop, it's remarkably precise and easy to use (with experience of course). For ALT I 3d printed a large knob to replace the t-handle for extra gronk and finesse, all the fine adjustment is done with the larger knob.
I routinely get way under 1 arc min PA on the EQ6R at least according to NINA. I guess NINA could be lying to me?
I always run the routine several times as a measurement is not accurate if it's not repeatable. I have confirmed PA with PHD2 several times and get <10 secs agreement.
Now whether one needs this level of accuracy for PA is another matter
