I saw 'The Mikado' at a local theater in December and it is fun to learn about a game and an agile process with the same name.
Often older software has design issues because of the data it was built to support. When exploring software that needs revising I frequently find that the data structure needs to be refactored as well. Does the Mikado method have strategies that support revising data and code together and testing both revisions?
The Method does not really care whether the prerequisites to a change are in the code or the data. It does however explicitly say that you should add /any/ prerequisite to a proposed change to your graph.
When it comes to data(base) restructuring the Method itself is agnostic but we believe that fast feedback from automated tests helps in those cases. And if you are restructuring data you probably want the tests to be on 'a different level' than the change itself.