I had someone from the Beeb prepping an interview on the Japanese solar sail probe last week who kept calling it a "space shuttle"., apparently under the impression that that was a general term for anything that went into space. Sigh. I propose the following correct astronomical and astronautical senses of 'satellite':

1) Any object in closed orbit around another object of larger mass (the most general sense, considered a loose usage: "the Earth is a satellite of the Sun" is rare, although "The Ikaros probe is a satellite of the Sun" does crop up. By 'closed' orbit I am excluding hyperbolic orbits - Voyager 2 was not a satellite of Saturn when it flew past.)
2) A natural celestial body in closed orbit around a nonstellar object of larger mass; a "natural satellite": "Phobos is a satellite of Mars". All known examples to date are rocky bodies, but one could imagine a Neptune orbiting a super-Jovian... the boundary between 'satellite and primary' and 'binary world' is fuzzy, as has been pointed out in the case of the Earth-Moon and Pluto-Charon systems.
3) An artificial object in closed orbit around any larger mass body: an "artificial satellite". "Space Shuttle Atlantis is an artificial satellite; the ISS is the largest artifiical Earth satellite; Cassini is an artificial satellite of Saturn". "The spacewalker's tool bag floated off and is now a separate artificial satellite" - so this includes all space debris objects. ("orbit" here implies gravitationally dominated motion: when Atlantis makes a flyaround of the ISS, it is not a satellite of the ISS. But possibly Luke's X-wing fighter, if his engines go out, is a satellite of the Death Star, even though the Death Star is artificial....)
4) An artificial satellite payload. "The satellite separated from the launch vehicle final stage". This is a narrower sense - satellite with a functionally useful payload as opposed to inert orbiting object.
5) A functioning artificial satellite payload. "How many satellites are there orbiting the Earth right now?". Often the questioner just means the ones that are still working.
6) An artificial satellite payload that does not include design provisions for carrying humans (i.e. is not a 'spaceship'), propulsion intended to send it onto a hyperbolic orbit after a brief stay in parking orbit (i.e. is not a 'space probe'), or aerosurfaces intended to provide controlled reentry and landing (i.e. is not a 'spaceplane'). This even narrower sense is the one being used by the original poster: a 'satellite' is an 'ordinary' spacecraft that isn't in any of these more interesting categories (but for some reason, other interesting categories: satellites with tethers for example, don't matter...).

I encounter frequent confusion caused by people mixing senses 3,4 and 5. Especially when they are asking me questions along the lines of the one in 5.
The X-37 clearly meets the definitions in senses 1, 3, 4 and 5, and so is a satellite *in those senses* even though one can argue that it's not *simply* a satellite per sense 6.. I, however, argue that sense 6, while valid nontechnical English by moderately widespread usage, should be eschewed by readers of slashdot as too muddily defined.

Exactly. Actually NASA TV has been dumbed down too much already compared to how it was in the late 1980s. The commentators speak over space-to-ground comms while repeating the same limited statistics
they've said 5 times already. We're geeks, we want data, so give us some different numbers - delta-V of the latest burn, what's the airlock pressure now, not just the official
landing time that the reporter was complaining about but the latitude and longitude of the landing site as well. That's the background info we need so we can go off and write the purple prose that Ferrell is looking for.

Better yet, stream the raw MOCR console data to us so we can crunch the numbers ourselves :-)

(Why yes, I am a scientist too. Why do you keep asking that?)