See http://en.wikipedia.org/wiki/Extremely_high_frequency.

Ted Rappaport gave an excellent keynote on this topic today at ICC2014 in Sydney. We will probably be putting the keynotes online in the next week or so; once it is available you will be able to get it from here:

http://icc2014.ieee-icc.org/speakers.html#Rappaport

The short form is that it has been experimentally shown to work very well in urban environments, contrary to what one might expect.

Yes, mm waves are the only way to offer considerable traffic in wireless communications. It is only a question of the approach to implementation.

4G networks are merely a spin and nothing else. There is no way of convincing lower microwave to deliver true 4G throughputs in real life conditions. Speaking of 5G in lower microwave is even a bigger lie.

Squeezing more bits per Hz as done in lower microwave is energy inefficient way of communication. A cell radius in urban environments have dropped to under 200m which came to the doorstep of mm waves, so yes, mm waves are a logical step forward.

While it is true that mm waves suffer from effects related to high frequencies, there are many other effects that compensate. One of the effects that vanishes with frequency is attenuation related to the ground surface. For orientational purposes, you may observe propagation attenuation at HF ~40log(distance) that goes down to ~34log(distance) at VHF, and so forth saimply because your practical mobile unit is close to the ground. In free space you have 20log(distance), and it nicely describes what you have in mm wave bands.

Absorption in gasses is detrimental for point-to-point links, but is beneficial for networked point-to-multipoint configurations, as it improves frequency reuse. At 60GHz you have virtually no need to make any kind of frequency planning if your cell radius is 200m or so. Point is that attenuation in gasses is linear per distance, which cuts your cell size to a sensible distance, and interferences from other cells is kept at minimum.

While high antenna gain is largely meaningless in point-to-multipoint networks, MIMO makes a lot more sense, and compared to low microwave a complex MIMO takes little space, and in effect you get the system gain equivalent to the antenna system complexity anyway.

If you observe nowadays mobile networks, and recognise that for the sake of information capacity - the cells are shrinking. It is easy to compare a mm wave system with small cells (limited by coverage) being the same size as low microwave cells (limited by capacity).

Yes, traffic congestion can be avoided only with higher BWs. Millimeter wave band offers almost  270 GHz,only thing we need take care of its atmoshperic attenuation