My suspicion, based upon watching the data, is that when present, it is usually uncorrected data, and when absent, the data has been corrected. I say this mostly because I normally see the 200nm+ circles near the start of a new day, and they tend to disappear as the day progresses, exactly the opposite from what you would expect for normally accumulating data.
Yup – when I looked at a few of those, I spotted things like 2100nm range at 2750ft altitude – not geometrically possible… When I shifted to running dump1090-mobility (shameless plug for obj’s work), a lot of those weirdies in the outer circle disappeared, but the overall numbers improved because of the way he does corrections and filtering.
I caught a plane in the act of creating the bogus data. The pattern was very interesting. I guess we can’t attach pictures here. If I find where I put it, I’ll upload it somewhere and post a link. Essentially, the points alternated between good points and bad points. The bad points weren’t random, though, so the overall pattern looked somewhat like an orb spider web at the scaffold stage.
You’ll get this effect if it has a flaky GPS lock for some reason - it’ll switch from GPS data to INS data.
(In theory, there is information about this in the ADS-B messages - they include the position’s uncertainty. In practice it’s tricky to work out what to keep and what to discard, because the largest categories are huge, but lots of aircraft send that anyway)
I am not a native english speaker so I may wrong on this but changing the wording to what you just gave makes it very clear where the data is coming from, while the “trailing” one does not sound so clear.