I use selfie expression almost exclusively and recently I started using my iPhone with ARKit, iFacialMocap, and VRCFT to have my face come more alive.
Through testing I have noticed that the face tracking with these items functions as intended, however when I enable selfie expression, it has no option to disable it's facial animations, causing a lot of extra shape keys to trigger and over stretch the face.
As of right now it is not appealing to use face tracking with selfie expression because there is no option in the menu to disable eye and mouth tracking with selfie expression.
Please consider adding these menu options in a future update. Thank you.
Edit: Example of broken shape keys shown in attached video.