@ubernaut We're not.
A) a number of the tunings we applied were needed anyway. They were recommended by the NodeBB people in a "scaling guide" and should have those in.
B) Going through cloudflare at this juncture is extra effort. One time largely, but we have something that works.
C) Likely would not have helped too much actually. The scraper used a bogus user agent and did not declare itself as a scraper. They were sending us about 2 requests per IP per minute. Individually, it was a tiny amount of traffic per IP.
(C) is a bit nuanced. Out of the box there is little for cloudflare to do on this one. There's some toggles I'm aware of though where you can have it help you defend against excess load. In that case, there could be help. But again, there's some tunings we should have had applied already.