Whirly
OK, we need to separate two separate issues here:
1) Why does the coriolis force exist (and what does it do) when we try to consider dynamics in a reference frame that spins with the earth?
2) Given that it does exist, how do we get from there to winds blowing parallel to isobars?
I didn't comment much on 1. Unless you want it in vector algebra, which many find isn't very satisfying as an explanation, all I can suggest is that you keep playing with oranges and turntables.
What's important is that you discover out of that process, and I don't think you're convinced yet, is that coriolis force acts to the right -- not the east, not the west, not the north or south, but the right (in the northern hemishere). Because it only acts on moving bodies, there's always a "right". Try it with a paper disk on an anticlockwise turntable. Run your pencil at constant speed in a straight line outwards, inwards, in the direction of rotation, against the direction of rotation. Look at the pencil traces. They always bent to the right as you look from where you put the pencil down to where you lifted it.
Now let's look at 2, and my car-on-slope model. The slope is like the pressure gradient on the surface of the earth. Its contours are like the isobars.
If you were to look at the scene from a distant star, yes, the slope (or the pressure pattern) would rotate with the earth. The whole scene would look very complicated, and we'd never work out which way the wind went. So we pretend that the earth is standing still, but in order to do so we have to invent this fictitious force (actually forces -- there's centrifugal force in the same category too). We fix the earth/slope/pressure-pattern but have to remember that if anything moves in this artificially fixed frame of reference, it feels a force to the right-- not the east, not the west, not the north or south, but the right.
The only stable state of motion for the car to be in is with the top of the hill on its right. The gravity of the hill pulls it left, the coriolis force pulls it right. If it had the top of the hill on its left, both forces would be acting downwards -- they wouldn't balance, they'd add. It's just like the wind: the anticyclone (high pressure) is always on its right.
If the car starts to track slightly right (going uphill), it slows down, so the coriolis force decreases. Now it has the same gravity pulling it down the slope to the left, but a smaller force pulling it up the slope to the right. So it gets pulled to the left, back onto its original track.
Any better?