I'm having some problems now, again with the tie. It's a long problem to work out, and may be confusing to explain. First of all, when I start the local simulation to test the cloth, it, after somehow being pulled towards the flamingo's chest, wrinkles and collapses into a ball.
After some contemplation, I decided that perhaps the gravity was off. The tie's construction curves were moved and tilted to correspond to the flamingo's neck at the beginning of animation. Since the tie's "down" position was now slanted and mostly to the right, I decided to adjust it so that it approximated the global "down".
After fumbling with the gravity values a bit, I decided, this time, to move the tie's construction curves back to where I made them, and move and rotate the tie from there. So I did that, and one curious side effect was that the tie's resolution (how many triangles it was tesselated into) increased. Another was that the constraints that held the tie to the neck all become scrunched to one side of the tie, and didn't seem to work.
So, getting mildly annoyed, I added new constraints to hold the tie in place. Unfortunately, when I started the animation (the real animation, not the local simulation), the tie remained in place, not moving with the flamingo's neck. After messing around and wondering whether the problem was the constraints, the gravity, or the tie itself, I gradually gave up, and now posted this up. What should I do?