Since the demise of mobile TV, the word "multicast" has pretty much gone out of the wireless industry. The idea of increasing efficiency of networks, by stopping the same content being transferred multiple times has pretty much dissipated. Yes, there is some discussion of cacheing certain files at various points in the network, but it would still need to be transmitted multiple times over the radio.
Obviously, side-loading between devices is an option, either via Bluetooth or memory card, but it can be a pretty clunky experience, and obviously iPhones don't have card slots anyway. It's also not good for realtime sharing.
The LoKast application from NearVerse looks like it might have some interesting potential. Although the initial use case has been local sharing of music (eg promos at SXSW in Austin), there appears to be a more general option to use it as a way of "offloading" mobile broadband traffic *to other handsets*. I'm pretty convinced that peer-to-peer connectivity among mobile devices is going to be a disruptor, whether it's MiFi-type products or virtual hotspots like Joiku's.
The CEO's phrase "unifying carrier and short-range wireless networks into a holistic, optimized system" in the funding announcement is an interesting comment that is evocative of various forms of cellular/WiFi (or in this case cellular/Bluetooth) hybrids.
Automated local sharing of content has some interesting implications for the UK's new and much-hated Digital Economy Bill too.