We went down a very similar route, are you running a Master/Local setup?
We were, so we took the local, and failed everything over to the master and then rebuilt the local with No AP's as a MC attached to a MM cluster.
We then moved a couple of AP's over to test before bringing the service up as live.
If you aren't running a master/local then this wouldn't work for you, Im unsure on other setups as I only know ours...