Define terms for things a UA thinks are/aren't tracking, even if it's wrong.

[jyasskin]

#2 defined "navigational tracking" (and "link decoration") around what's actually being done with a navigation and its associated URL. However, user agents often can't be sure what's actually being done.

There's a bit of an analogy to theorems that are provable in ZFC but not in a weaker axiomatic system like Peano arithmetic. In this case, "X link is not tracking" might be provable by an audit, but not by the software that makes up a user agent.

If we want to have user agents enforce that users aren't tracked using navigations, we're almost certain to have both false positives and false negatives, so we should have a second term describing the set of things a UA is expected to identify as navigational tracking. Maybe "identified navigational tracking"?

We might also want a third term in order to define a range of things UAs are allowed to identify as tracking, a bit like we had mixed content and blockable mixed content.