Vim is powerful when you need to script positional-editing actions (wrt the Vim cursor/lines/window, etc.) Couple that with the capability to record & repeat keystrokes and you quickly eliminate edit repetitions.

You can shell out Perl to perform the job of returning the LaTeX markup text, while you have Vim insert the text with a user command or key map. Then, you can figure out where in your document you wish to place the image commentaries and tell Vim the lines you wish to act on, in one swoop.