In the smoke simulation settings, there is an option called smoke adaptive domain. It basically (as far as I can tell) just changes the domain's size at each frame to be just big enough to fit the entire smoke simulation inside. It can really save some baking time, especially at the beginning of the simulation when the smoke is first expanding from the emitter.
My question is, is there any reason not to use smoke adaptive domain? I personally can't think of any reason not to. But it is an option, and it's actually off by default, so I thought there may be some disadvantage[s].