Nice summary from Jeff on whether the tool leads the process, or vice-versa.
On his point:
Lesson: Sometimes it works to capture with dark screen and then, just when
things are getting really complex and scattered, light up the map.
...a variation of this would be that if it's not acceptable to map with the projector on (or you don't feel ready to yet), you can still map privately during or even after meetings from notes (or pen+paper maps). Then in the reflective, less pressured (?) space *between* meetings you might spot problems or issues that have gone undetected (e.g. inconsistent decisions or use of arguments; contradictory interpretations of a problem). You could then go back to colleagues and say that when mapping the project you detected these (maybe give a glimpse of Compendium, but maybe not until they're asking to see how you do it)
Simon