(note: not my video, just one I googled)
On a dark (new moon) night, I've almost always been able to spot "shooting stars" aka meteors within less than 20 minutes if I lie on my back and look at the sky.
The smoke is interesting though!
Could it be possible to spin the entire satellite and get intermittent coverage and reception?
Speaking of those other sensors, spinning the satellite would mean pointing all of them away from where they are supposed to be looking – not necessarily a net win.
Finally, trying to spin the satellite at anything approaching 360 RPM might end up dramatically shortening the mission .
EDIT: In case you do not feel like watching the video,  shows a Little Joe II rocket (https://en.wikipedia.org/wiki/Little_Joe_II) breaking up due to an excessive roll rate (around 56 RPM) (https://en.wikipedia.org/wiki/A-003).
A recent, quite a spectacular example of that was the Hitomi X-Ray observatory which went into uncontrolled spin due to a faulty inertial unit.
To give you a sense of scale, if you reduce Earth to the size of a basketball, low earth orbit (space station, starlink, etc) would be less than a quarter inch above the surface of the basketball. Geosynchronous orbit is a ring around the equator 26 inches above the surface of the basketball.
This is also why Starlink is incomparable to existing satellite internet solutions.
Edit: This is the satellite being talked about in the article http://stuffin.space/?intldes=1998-030A
According to this graph , you need about a third of the delta-v budget it took to get to GEO in order to get back into LEO.