Not to shoot down (pun intended) another idea, but here is the way the orbital mechanics work. You can't change the laws of physics...
If you add energy to the satellites orbit (in other words, increase its orbital speed relative to the Earth), the satellite will gain altitude but still be in orbit. The reverse is true as well - if you decrease the orbital speed, the satellite will fall to a lower (and smaller radius) orbit. Since the radius is smaller the orbit is actually faster in completing one orbit (even though you decreased orbital speed to get to the new orbit).
So, to "bounce it out of the orbit towards the sun", you would have to add enough energy (increase the speed) of the orbit to a point that the satellite was out of geostationary orbit and then continue until the orbit was out of the gravity well of the Earth (hyperbolic trajectory relative to the Earth). The problem with that is you are still in an orbit 1AU distant from the Sun (the same as Earth's orbit).
Therefore you would have to start decreasing your orbital energy relative to the Sun to a point that the lowest part of the orbit now starts decreasing and the satellite gets closer to the Sun. The closest we've gotten a spacecraft to the Sun is the Messenger probe about to go into orbit around Mercury and that took multiple planetary flybys to reduce speed. This is because of how much energy it takes to reduce orbital velocity. The Sun's gravity well is very very large but we are in an orbit already and therefore you have to null-out that orbital velocity before something would fall "towards the Sun."
I know you were trying to start a conversation with your comment - so how's that for starters from the physics side? When geostationary satellites have warp drives, then we can talk about aiming for the Sun!
Andy
Just to add confusion, two things. First, readers might look at Escape velocity - Wikipedia, the free encyclopedia@@AMEPARAM@@/wiki/File:Newton_Cannon.svg" class="image"><img alt="" src="http://upload.wikimedia.org/wikipedia/commons/thumb/7/73/Newton_Cannon.svg/220px-Newton_Cannon.svg.png"@@AMEPARAM@@commons/thumb/7/73/Newton_Cannon.svg/220px-Newton_Cannon.svg.png which discusses escape velocity. The main implication here is that I think the discussion above assume that the added energy was tangential to the orbit, ie in the direction of travel, and that for small additions of energy, you are just turning a circular orbit into an elliptical orbit, and you're still in orbit. Escape velocity is usually thought of in terms of being directly away from the earth, ie perpendicular to the orbital motion. While I understand what is being said above, I intuitively think that you can attain a perpindicular component of the velocity that you can escape, orbit, so that in reality you are no longer in orbit. The above url refers to escape using a parallel burn having a parabolic shape rather than elliptical shape. Of course it's been decades since I've had to think about such things.
But the real reason for my reply here, with the partial intention of adding confusion , only because adding confusion often makes us think, is that while I think the above is worded properly, however I'm not sure that the reader gets the implications, not with respect to going to the sun, but with respect to the
part. This is often used as a discussion starter because it seems logically backward in that by doing a parallel burn to slow an orbital object, you are really increasing it's speed. Or said another way, if a low earth orbit object like the shuttle decays in orbit due to friction with whatever atmosphere might be up there, the friction slowing it actually speeds it up, because it's lowering it's orbit. Ie the actual speed of an object that has been lowered into a lower orbit is faster. For example, a TVRO sat is going about 6860 mph, while the ISS that's only 230 miles out is going about 17,600 mph. So when you slow these things down, they actually speed up.if you decrease the orbital speed, the satellite will fall to a lower (and smaller radius) orbit. Since the radius is smaller the orbit is actually faster in completing one orbit (even though you decreased orbital speed to get to the new orbit).
Actually, if NASA wants to decrease the orbital radius, it requires two burns, one to slow the ship down. This causes the ship to go into an elliptical orbit, ie when it slows down, it starts to fall a bit towards earth, but then speeds up due to gravity's acceleration, and when it gets to the lower protion of the elliptical orbit it has enough speed to bring it back out to the original point where the burn took place. To get the ship into a lower circular orbit, they need to do a second burn down when it is at the lowest point of the ellipse, the second burn again decreasing the speed to the point where the ship has the velocity associated with a circular orbit, thus two burns to slow the ship down have actually sped up the ship. While the 2 burn thing makes sense, it's still a mind bender (for me at least) relative to how a natural decay due to friction works, but I THINK that it's pretty much the same principle, since the friction slows the ship down so that the orbit's a bit elliptical, and then slows it even more at the lower part of the ellipse, keeping the orbit near circular as the friction lowers the orbit and speeds up.
Anyway, sorry for getting off topic once again, but this bumping the sat out of orbit thing reminded me of the friction speeds up the sat quandry.