Here are some best practice tips as well as some advice on what to expect when using Morph Cut in the real world.
Users must be selective when using Morph Cut and understand when it does not work in an ideal way (similar to using the Photoshop Content-Aware Fill feature). It works best with footage that features the following:
You can still get good results if all three of these conditions are not met. However, you can run into blotchy results, such as when there is much background, hand, or body movement. Face detection can also be blotchy when the subject is tightly framed or in profile. Even with those three conditions met, results vary. Face detection depends on things like:
Expect Morph Cut to work well when conditions are right. However, when working with less than ideal footage, Morph Cut is a less effective solution to heal a jump cut. Users who often work with interview footage can discern what works best for Morph Cut. Ideally, these users are setting up shots and subjects to be more conducive to working with Morph Cut.
Additional tips and tricks:
Morph Cut is a processor-heavy effect. Using it with large-format media causes slower analysis times, especially on GPUs with smaller memories (less than 2 GB VRAM).
Morph Cut has some inherent limitations. Problems can occur with movement in the background, far-away subjects, hands in the shot, or excess head movement. It usually takes trial and error to learn when to use it rather than reverting to dissolves and cutaways to cover a jump cut.
To leave feedback about the effect, create a feature request here.
For more information, see this post on the Premiere Pro team blog.