The downside, of course, is that your nuclear waste spends an awfully long time in an orbit that crosses, or at least passes very close to, Earth's. Better not lose control of your rocket, or you'll have a politically inconvenient situation on your hands.
the burn to fall into the sun happens way before any possible return to earth if that orbit were not altered... all that happens is that you start near the earth - thats unavoidable because the rocket comes form the earth.
it will of course cross the earth's orbit on the way into the sun, but, with the exception of very small ranges of orbits the chance of hitting the earth on return is tiny, and easily avoided with some timing.
I'm aware of how a bielliptic transfer orbit works. See the long gap between burns 1 and 2 on the Wikipedia diagram? That interval is necessarily going to be more than half a year long, and potentially much, much longer. Spacecraft are pretty reliable once they're in space, but that's still a lot of time for something to malfunction. And if burn 2 doesn't happen, now you have tons of nuclear waste on an elliptical orbit that re-visits Earth's path every so often. At any given perihelion, the earth has only a small chance of being in the right part of its orbit, but eventually...
I'm not saying it's a big risk; the chance of a catastrophic launch malfunction is probably a lot greater. But it's definitely a long-term potential hazard that needs to be accounted for.
unless burn 1 has a very specific effect that thing is not going to return to earth, despite crossing the orbit. hitting a resonance by accident is pretty unlikely... one could say astronomically so :P
