Oh, this is interesting.
I made a cute girl out of a slice of pizza using faceapp and headshot, about three weeks ago. Really just because I thought the idea was hilarious.
Your problem here is depth targets, which Headshot needs to calculate the contours of the face. Image is flat, so the AI won't know, on its own, how to capture that. But...
there are ways around that.Okay, let's see if we can composite our way out of 2d space. We start with this.

Then we take him into a compositing app, where we can take photos of people, and morph them together, to give him the depth targets that Headshot needs to work.
The big feature I want is the cheekbones, but if can get a good strong chin, that would be a plus too. Faceapp gives me two composites, which I can morph together.

And...

Anyway, it's a process, mashing pictures together, but all you have to do is keep repeating it until your photo is aged up enough, and all of his features look about right? It's really tough to do anything but guess on a cartoon character, because your brain needs the same depth targets as Headshot does. The sweet spot for older characters is about 40. Then, in cc3/headshot, you age them up by selecting older character, and you press the button. And, you get this. Well, that's not entirely true. I moved his nose, and scaled his eyes up a little. But yeah. Basically this