You might want to skip this longish introduction and go straight to the solution.

The problem
Mobile telephone manufacturers are slowly adding wifi to there telephones. Everyone seems to realize that the future is in IP networking. There is only one group strongly opposing this move towards wifi, the telephone service providers. Vodafone & Co don't like the idea of people talking to each other, without paying money for telephone services.
As the service providers are by far the biggest buyers of telephones they have managed to hold this technology back for a while, but sooner or later it will break through.
At that time the telephone market will change significantly, if it will continue to exists at all. The telephone service providers will probably be reduced to wireless internet providers, as any IP connection will do for VOIP. Their only advantage will be their existing antennae network. People will use WIFI to connect to their PC, and call from their, and only fall back the "telephone" network if their personal WIFI network is not available.

Another technology that will break through "any time now" is mesh networking. In a mesh network is a network in which hosts that can not directly talk to each other can use other hosts as intermediaries. In essence this is also the basic idea behind the internet. What makes a mesh network different is that the network is created ad-hoc, without preprogrammed routes.
The US Army uses mesh telephones in Iraq. If such a telephone can not reach the normal network, it will look for other telephones in the neighborhood, and ask those to forward the message.
The big disadvantage of mesh telephones is that networking takes a lot of power. To do it properly, all telephones must be on at all times. And the further a signal has to reach, the more power it takes. This means that the technology only works well if there are many devices around (this principle is known as Metcalfe's Law ).
In summary, for a good mesh network you want many devices at a short range, so you need only little power for transmissions.

The solution
My idea is to add a WIFI relay to every adapter. Everyone who has a telephone also owns an adapter for it. For most people this adapter is always plugged in, and almost always in the vicinity. There are so many telephones (and thus adapters) around, that any urban environment would be covered pretty well. Good coverage means that your telephone can get away with a weaker transmitter (less power usage).
Add a little mesh software, and a VOIP application, and suddenly all customers in range can call each other for free. Now that would be a good reason to buy a particular brand of telephone.

Infoworld reports that Motorola has decided that Linux will be their primary software platform from now one. New features will be developed on Linux first.
Recently I have been wondering on how Motorola thinks about Linux. Motorola is well on its way to replace Nokia as the biggest telephone manufacturer. As Motorola is also the frontrunner in using Linux on there phones, I wondered if there was any relation.
Either Motorola is becoming big because of Linux, or Linux usage is growing because of Motorola.
This announcement shows that Motorola considers Linux as a strategic advantage, and not just a piece of cheap software.
The decision to go with Linux probably also has something to do with Nokia being the primary backer of Symbian. It's hard to compete with the guys that also write the software to power your device. Although Symbian is more or less independent, it's clear that they are strongly influenced by Nokia. By not using Symbian Motorola can fight Nokia on two fronts.
Ofcourse Motorola could have used Windows Mobile, but that's a hornets nest in itself. Microsoft has an agenda of its own, and you never know how they will use their software against you. Once addicted to their products the quality falls rapidly (take a look at IE), and there is nothing you can do about it (except for replacing it). If there is a bug to fix, or a feature to add, you'll have to wait for Microsoft. In a market where a product is outdated after 3 months, that's unacceptable.
And if Microsoft doesn't like a feature (eg because it competes with a product of their own), you can entirely forget about it.

By choosing Linux the ball is back in Motorolas court. They can develop it as fast as they want, add all the features they need, and they don't even have to pay a dime for it. Ofcourse they will have to pay programmers to write the software, but if you are as big as Motorola, that's not a problem at all.