Congratulations, you discovered an important machine learning meta-algorithm: http://en.wikipedia.org/wiki/Weighted_Majority_Algorithm.

No, network coding is strictly better than multipath routing, which is simply simultaneous transmission along multiple routes. Multipath routing still looks at information networks as commodity flow networks. Please see the butterfly network example in the network coding article again in Wikipedia to understand why information flow networks are different and allow for higher capacities and pure routing cannot achieve this. Hope you will appreciate the counter-intuition.

The paper is not a breakthrough in terms of the concept itself, which has been around for quite a while. And the concept is truly remarkable -- as much as if 2 + 1 = 4 (which is essentially the case in the butterfly network). But practical coding schemes (in terms of computational complexity) and implementations are still works in progress.

Please see Network Coding: http://en.wikipedia.org/wiki/Linear_network_coding [wikipedia.org]. It is beyond simple FEC and can increase the capacity in networks over traditional packet routing schemes, counter to our intuition for flow networks/water pipes.

This is not simple data compression or error control coding. This is network coding, e.g., see http://en.wikipedia.org/wiki/Linear_network_coding [wikipedia.org] and how it can increase the capacity in the butterfly network over traditional packet routing schemes, counter to our intuition for flow networks/water pipes.

It is a fairly hot research topic that has been around for last few years. But it is fairly revolutionary. It is still early days in terms of practical coding schemes and implementations.

This is not plain FEC for point-to-point communication. This is based on network coding, e.g., see http://en.wikipedia.org/wiki/Linear_network_coding and how it can increase the capacity in the butterfly network over traditional packet routing schemes, counter to our intuition for flow networks/water pipes.

Network coding has been a fairly hot research topic in information theory and coding theory over last few years. But it is fairly revolutionary in my opinion. It is still early days in terms of practical coding schemes and implementations.