How much of that is "New and Improved!" with the emphasis on improved, as in, better/cheaper ways to do the same thing, vs. just plain new? I cite the FPGA because while it was the end of a long line of innovation like PALs, it was qualitatively different. E.g. allowing telecom vendors to support new or improved ^_^ protocols with new firmware (vs. an inflexible ASIC or the like or very very fast general purpose CPUs).

The cloud, yes (and a classic disruptive innovation). SDN and virtualization, I'm not so sure. Or at least the latter was just applying old concepts (it's been a long time since IBM's main OS could boot on raw hardware) to a different platform. Radio I don't follow at that level. Low power/high performance eventually equals a qualitative difference which I gather we've reached, so yes, but I quibble that was semi-inevitable, driven by demand and specifically enabled by Moore's Law and the savvy of ARM Holdings et. al. Camera sensors even more of that, plus also enabled by low power infrastructure.

ADDED: no breakthroughs in batteries, but some incremental improvement?