Feels like the "right" approach would be to have it write some code to measure how far off the elements are in the original vs recreated image, and then iterate

Feels like the "right" approach would be to have it write some code to measure how far off the elements are in the original vs recreated image, and then iterate using the numerical output of the program...