An ability to move single components around the picture space and have the picture behind fill in automatically. or even manipulation of specific item without comprimising the resto of the picture. Like to centre the person or move the person into more of a sprint form. I mean it might already work that way I don't have a lot of experience with text to image yet.