I regularly criticise the notion of mobile centrex as a plausible standalone replacement for PBXs. Absolutely, it has a role to play in certain niches - salesforces, certain small companies (if you can get the channels right) and so on. And in the context of a larger consultancy-driven managed services offerings (think BT Global Services), with plenty of IT systems integration involvement, there is a possibility of more wide-scale adoption of telephony "in the network".
But I still keep getting blustery press releases from vendors and operators which essentially say "throw your PBX in the bin on Friday night, and start using your mobile handset instead on Monday morning, with your operator as your service provider".
I've posted at length before about why this is utterly unrealistic as a massmarket, generic proposition, whether it is FMC-based, IMS-anchored, IN-resident or any other variety.
So I've thought up a simple test for whether this can fly in reality, particularly in a heavy-duty, mission-critical enterprise context:
Show me a mobile operator which runs its own main customer call centre via a network-hosted platform, rather than a discrete ACD or PBX. And which has taken away all the deskphones from its finance department. Then (and assuming it works!) I'll start to change my mind.