In Nima's lectures on the 'Twistor Uprising' (for example here), he gushes about about new powerful techniques for calculating amplitudes, such as summing Feynman diagrams using 'BCFW recursion.' He makes it sound like there should be practical results, like new QCD amplitudes for use in Monte Carlo generators, etc. Are there any, or are any on the horizon? Or does he overstate the power of these new techniques?
EDIT: There is a recent blog post by Matt Strassler seeming to imply that BlackHat (a program that computes QCD amplitudes) uses some of these techniques.