It's also now possible to do clinically-useful testing of targeted treatments using genetically-edited control tissue cultures against unedited tissue cultures.
The point is that it's better to test something that is as close to the patient as possible, not guess with other species or even other people whom express genes differently.
FYI: making a new mouse models requires lots of chopping off heads of mice whom don't possess the desired gene. While transgenic, highly-edited living models might be nice, there's probably a more humane/simpler/reproducible way to do the exact same thing.