Would editable layers make AI-generated figures actually usable?

Many AI image tools generate fast, but editing is painful.
In ScholarViz, we decompose generated scientific figures into editable text/object/background layers so users can move, restyle, and fix details precisely.
Would this be useful in your workflow (papers, posters, slides)?
What’s the #1 layer operation you need most?

https://www.producthunt.com/products/scholarviz?launch=scholarviz