Los Angeles Noir, a 2011 film made by Rockstar, surprised many with facial animations that were far superior to other games. The technology used at this time is called MotionScan. Basically, the cost is very high because the actor sits in a room where several cameras are elaborately arranged to perform, capture and reprocess it. So, it's also a method that you can't try unless you're a big game company.
The “First Order Motion Model for Image Animation” released in 2019 uses deep learning technology to create virtual character facial expressions, which has the advantage of dramatically lowering the required cost rather than enabling something that was not possible. All you need to prepare is one camera and input images. There is a little more work to be done to apply it to the actual game (applying 3D models, generating high-quality results, candidates, etc.), but it will soon become a technology that has no reason to not use it.