At risk of sounding like a prediction, we can confidently say the outcome is something awful. The ecosystem isn't built like a fault-tolerant network where one node goes out, but the network more or less still works. There is no redundancy. There are no enterprise-grade availability requirements.