[kristofferR]

It has been corrected to be $11B over a five year period, not annually as it was originally stated.

Considering that Google almost hit $11B in profit in 2012 alone, spending $11B over 5 years on a revolutionary infrastructure project that might even start making a direct profit directly from the monthly fees after a while for itself, but most importantly could totally revolutionize what type of services and experiences Google could deliver to ~15% of American households - doesn't seem like such a bad investment after all.