"Focus" is an inconvenient word if you're thinking of changing the potential, because if you do then the orbits are no longer conics and the word kind of loses its meaning. That aside, let me see if I understood your question correctly:
Given a gravitational potential that's spherically symmetric around a central point $\mathbf{r}_0$, and which has a gravitational potential $V(|\mathbf{r}-\mathbf{r}_0|)$, what's the fundamental reason that orbital speeds decrease as $|\mathbf{r}-\mathbf{r}_0|$ increases? Is this due to the gravitational field being weaker at longer distances?
In that case, the answer is that orbital speeds decrease because $V$ itself increases at longer distances. This is simply conservation of energy:
$$\frac12m\mathbf{v}^2+mV(|\mathbf{r}-\mathbf{r}_0|)=E,$$
and if $V$ becomes less negative then $v^2$ must be smaller. Thus, potentials where this doesn't happen must have regions where the potential is repulsive from the origin. One such example is
$$V(r)=-\frac1r -r,$$
though of course there's no physical system with that behaviour.
You've asked a very entertaining question, and the answer is not simple.
Let's ignore collisions for the moment. The "purest" effect, that is, the one which involves no change on the part of the planet or its sun, is the effect of tidal bulges in the sun. Just as the earth, for instance, is not a perfect sphere due to tidal forces, so the sun is not a perfect sphere, due to tidal forces caused by the earth. The resulting bulge in the sun lags behind the planet, and essentially acts as a brake on the planet. Over time, the planet will gradually lose velocity, and will eventually fall into the star. For most planetary systems, the effect will take a very, very, long time, since the planet is much smaller than the sun, and far away.
But there's another factor to consider. Any star produces a "solar wind" which causes it to lose mass. The amount lost per year is small, but it never stops. The result is that, over billions of years the planet's orbit will grow larger as the gravitational attraction to the sun diminishes.
Finally, for stars like our sun, stellar evolution will eventually cause the star to become a red giant. If the diameter of the star exceeds the orbital distance of the planet, of course, the planet will be vaporized. Even it if doesn't, the tidal bulge will become much more effective in slowing the planet, and depending on details of the planet's orbit may or may not cause the planet to drop into the star before the star shrinks to red dwarf status.
In the case of the earth, according to http://arxiv.org/pdf/0801.4031v1.pdf that is exactly what will happen to the earth in (roughly) 7.59 billion years. It's notable that if the earth's orbit were 15% larger it would be safe. But just before the sun reaches peak diameter tidal forces conspire to slow the earth down and it plunges (will plunge)into the sun.
As for other considerations, such as explosions, impacts and shock waves, the answer is that they can have an effect, but generally not much. Basically, if the impact or whatever were big enough to make a major change in the planet's orbit, the planet would cease to exist, and would be replaced by a great big debris field. To some degree this would recondense into a smaller planet with a different orbit, but it wouldn't be the original one. Just as a thought experiment, though, if the earth were to hit another earth-sized body exactly head on, and the other body were in an identical orbit but going the other way, and the two planets fused instead of turning into a massive debris field, the resulting fused body would drop straight into the sun.
As for a planet ageing, for earth-types the answer is, not much. It's true that our kind of planet can lose volatiles such as water and air (and do so at a very low rate), but the total effect is miniscule. We are, after all, mostly rock and iron, and those just don't go anywhere. For gas giants like Jupiter, if they are close in they can get their gasses blown off until there is nothing left, or only the non-gas core. However, any such loss will be at right angles to the orbital motion (for more-or-less circular orbits) and will have virtually no effect the orbital motion of the planet.
Best Answer
Assume a high circular orbit above a planet. If you want to drop to a lower orbit, you have to do a small retro "burn" (fire your rocket engine to provide thrust opposite the direction that you are traveling) to reduce your tangential velocity a bit. If you don't slow down too much, you will go into an elliptical orbit, gain kinetic energy as you drop lower in altitude due to a decrease in gravitational potential energy, and approach the lowest point in the new orbit at high speed. The speed at lowest approach (perigee) will be too high to remain at that distance from the planet, and you will eventually rise back up to the point where you fired the retro rocket. To prevent this, you have to do another retro burn at perigee to go into a circular orbit. Once this happens, the orbital speed will be higher than it was at a higher altitude, with that increase in speed coming from the decrease in gravitational potential energy minus the decrease in speed from the second retro burn.