Because, like most teams, we had certain resource constraints (time, money, datacenter space, employees, etc). The undertaking of a massive new project that puts greater-than-ever reliability demands on existing infrastructure that is properly sized for the existing tasks is a huge burden.