That's sort of the point, though. Is there anything in the "do not track" policies that prevents the data owner (Reddit in this scenario) from just selling its own scraped data on the end-user to an ad network? Ad tracking is functionally equivalent; the tracking scripts just automate and simplify the process.

"Do not track" is a bit of a toothless concept when the very nature of the medium is to make auditable connections to remote servers.