this question is actualy somewhat erronious, it assumes that naturaly occuring orbits are somehow unusualy stable, but thats purely because all the unstable orbits decayed or caused collisions billions/millions of years ago, and as we know from places like Barringer crater or even Tungsta (maybe), there are still objects in the solar system with unstable orbits, so to assume that natural orbits are somehow blessed with some extra stability is in error.<br /><br />with our modern equipment we can detect peturbations in orbits down to miniscule amounts that make the natural drift of (say) Earth look crude and innacurate, if we can detect errors in timing that apear to our instrumentality to be gross then it is obviously possible that we could fine tune the orbit of an artificial object to a higher precision than anything else in the solar system, so we can surpass any natural orbit easily, and probably have with something like Gravity probe A.<br /><br />Geostationary com satelites such as those used for TV etc tend to get pulled out of their designated places by the variations in gravity as the month progresses and the moon tries to adjust their orbit, that plus air drag (V slight but there), the sats have to be in the right point of the sky cos thats where your dish is aimed, so they need to stationkeep within 30 miles or so of their intended position, plus if the satellite turned by one degree the microwave horn feeding the sports channel to your TV dish would be pointed at the wrong spot on the earth (they all have tightly controled "footprints"), so due to air resistance and lunar/solar pull (mainly) the operators need to keep em on station with reasonable precision with the odd manual "tweak"<br /><br />all this tweaking is nothing to do with the innacurracy of the satellites orbit, but due to the operators wanting it to follow an un-natural orbit that is unaffected by anything else in the solar system (and atmospheric drag). so that you can use a fixed dish to point