anyway, how likely is it the ice ages could be explained by the earth
'realigning' so that polar regions would migrate over the surface of
the earth?
How about zero? The geological evidence of the Ice Ages clearly says that, between the ice episodes, the ice did not move. It's just that the polar caps shrank. For instance, the extent of the last ice age is well mapped, and it is centered on the North Pole. It's just that the ice cap got a whole lot bigger as the planet got colder. When things warmed up the ice sheets retreated, but the center stayed at the North Pole.
While much older ice ages left their marks at different places on the current globe, you must keep in mind that, for the entire history of the earth, the continents have very slowly been sliding around, so (for instance) Antarctica used to be much farther north than it is now, and was correspondingly warmer.
As for
It appears to me all the earth would need is a small 'nudge' and it's
existing angular momentum would send it toppling end-over end
well, no. Just no. The forces required would be enormously more than the sun CAN apply by any mechanism that we know about. A small nudge just won't do it. The earth's rotation is not perfectly static (see the precession of the poles) but neither is it grossly unstable.
EDIT - It has been suggested that the anomalous rotations of Venus and Uranus might be examples of such a flip.
The case of Venus is not well-understood, and there are at least two different possibilities: peculiar formation due to perturbation during coalescence and tidal effects. In either case, the theories do not predict a flip on the timescale of 50,000 years. In the first case, no flip of the finished planet occurred, and in the second case the process took in excess of a billion years.
Uranus is even less useful, since it apparently occurred as a result of multiple impacts. On the one hand, the axial inclination is 98 degrees, so the planet is essentially lying on its side (compared to the rest of the solar system). More importantly, its satellites orbit in a plane which coincides with the rotational axis. This means that a shift in the orientation of the planetary axis cannot explain its current position, since such a shift would not have affected the orbits of its satellites.
END EDIT
What may be confusing you is that the passage you quote about the Space Shuttle is talking about speed relative to a fixed frame of reference: one fixed relative to the distant stars. On the other hand, when you think about aircraft flying through the air (or people walking along the ground), you think about a co-rotating frame of reference: fixed relative to the Earth itself.
Each of these frames of reference makes sense on its own, but mixing the two together makes a mess.
In the fixed frame of reference, the Shuttle needs to orbit at 17,000mph. This speed is the same whether it is orbiting west, east, north, or south. In this frame of reference, the Earth is rotating at 915mph eastwards - which also means that the Shuttle, just before take-off, is moving at 915mph eastwards. Consequently "915mph eastwards to 17,000mph eastwards" requires less effort than "-915mph westwards to 17,000mph westwards".
In the co-rotating frame of reference, the Shuttle needs to orbit at 16,085mph if it is orbiting eastwards or at 17,915mph if it is orbiting westwards. In this frame of reference, the Earth is stationary, and so is the Shuttle just before take-off. "Stationary to 16,085mph" requires less effort than "stationary to 17,915mph".
The fixed frame of reference makes more sense for spacecraft because it makes all orbital speeds the same.
For aircraft, both frames of reference are again possible, but in this case the co-rotating frame of reference makes more sense because aircraft travel through the air, and the air rotates along with the Earth. For simplicity let's fly along the Equator.
In the co-rotating frame of reference, the aircraft flies at 560mph eastwards or westwards, above a stationary Earth. To get to a destination 560 miles away, it flies for an hour.
In the fixed frame of reference, the eastbound aircraft flies at 1,475mph eastward, above the Earth, which is rotating eastward at 915mph. After an hour, the surface of the Earth has moved 915 miles, so the aircraft is $1475-915=560$ miles ahead of it. The westbound aircraft flies at $915-560=355$mph eastward, above the Earth, which is rotating eastward at 915mph. After an hour, the surface of the Earth has moved 915 miles east, the aircraft has moved 355 miles east, so the aircraft is above the point on the Earth's surface which is $355-915=-560$ miles east (in other words, 560 miles west) of the starting point.
What makes the air move at the same angular speed as the rotation of the Earth? Simply this: that if one layer were rotating faster or slower than the other, drag would speed up the slower layer and slow down the faster one.
Best Answer
The friction due to a single person would be on the order of ~100N. There are currently 7.2 billion persons on Earth. I don't think this will affect revolution around Sun but surely it will affect the rotation of Earth about its own axis.
I did calculated this effect on the back of envelope. Assuming a very not-so-realistic assumption that all people will march along the equator simultaneously, we can calculate an upper limit on the total external torque opposing the rotation of the earth. The angular acceleration due to such a march will be so small that it will take something on the order of 100,000,000 years for all the people to march with no stop to finally being able to bring the rotation to a halt. And, obviously no one can do this for such a long time. So, we are safe.