I treat them as sketches, the way Martin Fowler used to describe a UML diagram. They have to be either disposable or treated as a notepad. I reconstruct the full picture only when it's necessary from "sketches". E.g., since Lucid has an infinite canvas, I'll put all my different diagrams on a single canvas, along with little diagrammatic "ideas", "notes", "brain-f$rts", etc.

For a presentation or project documentation artifacts, I can then go thru my sketches and pull bigger pictures together easily. I applied Fowler's "sketch" principle in my own tool ormle.com