If you end up with a scenario there, though, let's talk ... I can definitely imagine some, especially for cascading updates to multiple stacks (e.g., "update compute resources", then "update applications" on top) ...