Think of anyone you’ve encountered whose native language is different than yours. Chances are they had an accent. The less fluent they were, or the more different their native language was to yours, the thicker the accent. The thicker their accent, the harder they were for you to understand.
What about the application to serial criminals, where you have one person committing a series of crimes over time. Wouldn’t geographic profiling and ML be better suited for this?
It depends on exactly what is learned from looking through them. If you end up copying shapes and segments then there are possible grounds for a lawsuit. If you’re able to determine the rules to make a good font from your analysis, however, then nothing is stopping you from applying them.