When discussing demisability in
yet another thread, I had a thought: Why aren't we (well, the Feds) requiring actual demisability tests for these LEO megaconstellations instead of just calculations and estimates? Apparently the estimates for the Dragon 2 trunk were... optimistic (~33 re-entries, debris found 4 separate times.) Conversely Starlink seems to have excellent demisability (~300 re-entries, debris found 0 times)
It wouldn't be that expensive to rideshare with a Starlink launch or something. Just need enough control to have the test satellite come down in a known area with good tracking/optics/etc - maybe some isolated missile range.