I'd reckon on Nokia needing something like four phones:
1. Small basic calls and text with a traditional numbered keypad
2. Basic touch screen device aimed at web browsing and social networking
3. "Business" or "Chat" device with a built in physical keyboard
4. Super Computer in a phone
They could iterate no 4 fairly rapidly, it seems to work well for HTC they're hardly ever out of the news with their rapid release cycle.
Say we get hard intel that sometime later that day, someone will be using Twitter or Gmail to issue timing commands to a bunch of people ready to drop off backpack bombs on metro trains in half a dozen large cities around the country. The "kill switch" mechanism doesn't shut down the internet. It allows the counter terror people to ask the administration to use that legal power to get on the phone with Twitter and tell them what needs to happen to prevent such use.
What exactly do you need to legislate for? If there was genuine intel that terrorists were going to use Twitter to arm a device or orchestrate an attack, do you really think that when presented with this information and a request to make changes if order to avert an national catastrophe that Twitter would turn around to the government and tell them to get lost? Even if the data was being transmitted on services outside the US and it required changes at a network infrastructure level to disrupt peering of data from those services, do you really think that a company would turn down a request to make such a change?
"If the code and the comments disagree, then both are probably wrong." -- Norm Schryer