To summarize:
I claim that diagramming software is useful for understanding and planning structural changes, but
current tools use a one-diagram-fits all approach without asking the programmer what they want to do to the program.
Dr Jones collaborates with the programmer in exploring the program’s design by understanding an economical vocabulary of refactorings.
By doing so, it can create focused, relevant and simple diagrams.
I haven’t wrote out a formal research plan, but my medium-term goals would be to put in writing my answers to challenges 1, 2, and 3 (as formally as is needed) while working on the implementation of Dr. Jones.