I've suggested this idea in an earlier FT post but you've since added more compatible and worthwhile technology like AI Headshot and the Wrinkle System so I am pointing out the existing tech you already have to make this possible.

A superior product catered to every unique users face for facial motion capture and wrinkle system would be:

1.Capture sound and facial expression from AccuFace - Since the green calibration dots recognize depth, then use the dot depth tracked data with-
2. AI in HeadShot 2 to wrap that depth information to create the initial face shape if needed or create the morphs and wrinkles. AI can also be used to estimate contours between a photo and the depth tracking dots -Adobe Substance 3D Sampler captures depth from color photos fairly accurately -even from a single photo!

3.You should be able to combine the depth tracking from Accuface calibration for a particular facial expression and use it to conform the mesh to a morph - no need to sculpt individual facial expressions - it's like using Headshot 2 for every facial expression and Visme. Like Adobe 3D Sampler, Depth estimation from a single photo may also help make this more accurate. The AI would be told to reform areas of the face based on differences rather than recreate the entire face to save time.

Highlights:

***Is a universal facial expression capture unique to the user's characters whether human or not saving days of sculpting time
***High accuracy of facial expressions matching the customer's face and would be even greater realism using a digital actor scan of the same person, the film industry would go nuts over this product for digital doubles.

***create several facial motion capture profiles for each actor to have the character's face better conform to facial animations
***Simultaneously capture/create facial morph and corresponding/matching Wrinkle -even more time saved.