Although Amdahl's law does limit the parallelization efforts, it only applies to fixed problem sizes. Greater parallelization may not always lead to greater performance of existing problems but instead leads to the ability to solve new problems that were impossible before (Gustafson's Law: http://en.wikipedia.org/wiki/Gustafsons_law).

There is also a great new area of research dedicated to building load balancing algorithms that not only help computing performance but also attempt to minimize energy. Using energy efficiently will be great limiting factor moving forward as our hunger for processing power continues to grow.

The title is kind of a bait. The article is an interesting-enough discussion of Amdahl's law and how it can limit efforts to parallelize stuff at google.