Indeed David.
Phantom is indeed common mode therefore so is its ripple, and anything common mode should be rejected easily by the preamp.
However...
at the Microphone (or a phantom-powered active DI box, or what-have-you) the mic is driven by the phantom power. The power to polarise the capsule is summed 'common-mode' and it is the audio output signal which cancels. Any ripple at the phantom supply will show up at the microphone... -whether or not the mic rejects it is altogether another matter.
Of course many microphones then filter significantly, or even use HF multiplier circuits followed by filtering, (thinking of sensitive tasks like polarisation voltage) and this will usually completely elinminate most noises, but there are most certainly some simpler designs out there which do indeed tend to 'pass on' any audible noises in the DC supply into their output signal, and at that point there's no longer any way to remove them.
-I used to dismiss phantom ripple rejection as a potential problem, because I always looked at it from the preamplifier's 'viewpoint'. -Once I started looking at it from the microphone's 'viewpoint', I realised that it probably is worthy of a little more effort to ensure that it's kept pretty quiet. (I'm not a nutcase about it though!)
I should probably also confess to the fact that it was an active DI which I designed and built which got me thinking about this matter... It had pretty poor supply noise rejection (it was small, simple and cheap!) -but none the less there are definately mics out there which are less able to reject supply ripple than others.
Keith