When I was at google, the one thing I took away was that most of the terrible decisions such as described above were very widely panned before being released inside the company. Everyone one of them. That doesn't mean the group of engineers was always right, but they were very good at pointing out bad, controversial ideas.

> cloud business is synergetic with Google search

The funny thing is that while it's obviously synergetic, up to 2012 it was a Urs' toy project on the back-burner. It was mostly run from SEA/KIR with MTV leadership largely oblivious. There were gems like this, for example: once they forgot to cover cloud organization in the company OKR meetings broadcasted to all employees :) Only when they realized that external clouds grow so fast that Big G can loose it's preferential access to volume discounts GCP suddenly became the priority (to detriment of TI, which now largely is after-thought).

I'm convinced that Google going to cloud too late is going to be seen in history books in the same light as Microsoft/IE fiasco.