I'm not so sure of this. Google has published multiple documents [1] mentioning the limitation of top-of-rack switching capacity and data center bisection bandwidth: how they need to design around it, and why it makes sense to use commodity switches despite the burden on software design.
I didn't read it to see what you were referring to, but that's the second edition of an older book, so maybe they didn't update everything for the new edition. Google has been banging the software-defined networking drum for a while now (configuration is part of the compute engine APIs now, for instance[1]).
[1] Here is one less than a year old: http://www.morganclaypool.com/doi/abs/10.2200/S00516ED2V01Y2...