I’ve encountered some good news this morning. The voice cloning tool recently introduced “labels” on your clones, allowing you to specify how you want the AI to produce the voice. I added in a slower pace and more gravelly voice variables, and this was the result (based on Burbin’s line suggestion):
https://drive.google.com/file/d/1Zo03jHaSIST71EEYlxogDI-TJNPWoEUj/view?usp=sharing
https://streamable.com/6xbykh

This sounds extremely similar to the way Mark speaks in the film. For a contrasting example, here is the same line with just the slower pace variable (no gravelly variable):
https://drive.google.com/file/d/1cBHFTojz0GmsCK8aRkKhS-GM92Xffplp/view?usp=sharing