It's kind of depressing that in 2011, when LTE is starting to roll out in some countries, and some 8 years after 3G started rolling out around the world, that the usefulness of 2G fallback is still even debated.

2G is a dead technology, and the sooner its buried the better. It can't do data and voice at the same time, (except for a grafted on, almost never supported extension called Dual Transfer Mode, which in the UK only Vodafone and a few new Nokia handsets support) maximum speeds and number of simultaneous users is pitifully low, and the encryption is well and truly broken, allowing for "man in the middle" attacks, eavesdropping, and spoofing using portable equipment such as that recently bought by the UK police.

2G is such a broken technology by todays standards that its as if we were all still using 1999 era WEP encryption on our Wifi, despite being fully aware that its completely broken. Cracking of WEP lead to widespread adoption of WPA and WPA2, yet cracking of GSM lead to.......nothing. GSM is still as vulnerable as the day it was cracked many years ago, and despite the fact that UMTS (without 2G fallback) fixes many (most?) of the security vulnerabilities in GSM, somehow fallback to GSM is still seen as a good thing.

People and networks need to wake up and smell the coffee, and stop running their businesses and lives on ancient, insecure, slow, and woefully inadequate technology when there is not one but two later generations of technology now available. (Is anyone still using or is satisfied with 56k dialup internet access at home ? Yet what we have on 2G is similar or worse than that)

We don't even have ubiquitous 3G coverage on low frequencies in the UK yet, and LTE is nearly upon us. 3G today is entry level, not high end, and it baffles me beyond belief that in this day and age networks cling to 2G so tenuously.

There really needs to be an aggressive push to both massively roll out low frequency 3G, start deploying LTE, and do the mobile phone equivalent of the analog tv switch-off - warn everyone that in x number of years 2G will be switched off for good and that they have that amount of time to upgrade to a 3G capable phone or lose service.

Perhaps it would take a mandate from Ofcom to do this across all the networks, and I doubt they have the spine to do it, but we can hope. Continuing to provide 2G services in the long term is a gross waste of bandwidth allocation - 3G allows a far greater aggregate data throughput and call throughput per cell for a given bandwidth allocation - so by keeping a large chunk of spectrum in 2G use maximum potential is lost, and there is less capacity available for end users.

Progressively re-farming existing 2G spectrum to 3G maximizes the use of the frequency allocations, and should be part of a progressive move towards a complete 2G switch-off.