Coriolis force is almost fictitious. It's quantative description in classical mechanics seems worse than in quantum mechanics. (I've derived and applied coriolis matrix elements of the Haliltonian of vibrating-rotating polyatomic molecules. Assymetric rotors are a bit nasty to apply it too, but that's because of the way one has to handle assymetry. In symmetric tops it's not too bad, but they do depend on the symmetry of the vibrations involved [In methane Coriolis interactions get mixed in with so may other things that after many man years simple vibrational band spectra can only be quantatively fit up to J=12 (as of 12 years ago). A. R. at Oxford worked on this for years, and did so many model calculations on Oxford's computers, that he became the expert on them and computer methods. Deciding that computer technology was yet to primative to tackle methane well, he accepted the offer of the Computer Science Department to head that department, and resigned his chemistry professorship.] There is no Coriolis interaction between two totally symmetic vibrations.

Goldstein's 'Classical Mechanics' treats Coriolis interaction fairly completely, and shows that the force of it is less then .001 time the force of gravity (pulling the water into the drain). The drain problem isn't the only one where local perturbations can easily completely overshadow Coriolis effects.

One neat Coriolis effect is setting a pendulum swinging at the north or south pole. The poles are moving around the sun, but don't rotate with the earth. The pendulum's motion with respect to the sun is practically spaced fixed. However, to an observer at the pole (who does rotate with the earth) it looks like the pendulum is precessing, with one revolution each day.

The omega cross velocity vectors that give Coriolis force is a bit messy, because they're in a space fixed coordinate system, and we observers aren't, and find the equations for motion in our local xyz and 3d rotional coordinates are quite messy.