TheGreendaleObservatory PHD2 and PEC do not mix - they fight the corrections and I've been an astrophotograper for many years .... I don't know one person who uses PEC due to that
If the PE correction is designed properly by the mount, so that it is additive, it can be made to work with autoguiding, but it has to be implemented to explicitly allow that. That is assuming that the gears are so well manufactured that one rotation is precisely the same as the next rotation. Most mounts, and it certainly includes the ZWO mount, does not have that kind of repeatability (less than 0.2 arc seconds) from cycle to cycle.
The reason autoguiding is not compatible with shaft encoders is that autoguiding will try to move the mount based on the star centroids, but the shaft encoder is moving it back base on the encoded position; thus "fighting one another."
But PEC is simply a "blind" curve that the autoguding corrections is added to, not fighting against, so it is not impossible to implement correctly so they complement one another instead of fighting one another, as with shaft encoders.
If implemented properly, PEC has the potential to reduce the amount of error autoguide has to correct, and perhaps produce a more precise guiding.
Since the first poster mentioned ASIAIR, I assume people are talkng about PEC for astrophotography, and not for visual work.
PEC, while useful for visual work (since your eyes can follow a target even if it is many arcseconds away; your eye-brain are basically "autoguiding"), it is usually not sufficiently precise (unless you are talking about very short focal lengths) for astrophotography purposes -- i.e., you will still need to autoguide (or buy a premium mount).
However, like @TheGreendaleObservatory, I never use PEC myself. I get enough precision though my autoguiding that I do not need to add PEC (another layer of complication that taxes my simple mind, and I do not own a premium mount).
Chen