Fascinating idea. I'm curious how you would be able to direct traffic to all of the spare computers and networks to absorb the attack. I guess I'll have to read the whitepaper.

On a side note, it would be great if you could source that F.U.D. you are creating in the following:

Cloudflare was DDoS’d and shut down for a few hours this summer.

Cloudflare was not taken down due to a DDoS attack. (Source: I work there.)
Are you thinking the Dyn or AWS outage?