That's not a simple question, as it depends on the relative stiffness of wall and fill, freq content of the EQ, etc. If the wall is "stiff" (such that its natural frequency is high with respect to that of the fill), the soil and ocean water pressure will be out of phase, such that the fill and its pore water - think in terms of undrained soil behavior because of cyclic loading with high frequency - are being left behind by the wall when the wall is pushing into the ocean. (That is, the hydrodynamic pressure is highest when the fill pressure, which would help resist it, is lowest.) It gets more complicated if the wall is higher and more than just the principal mode of vibration is excited, but that's probably not a big issue for most structures.
One question that occurred to be but I can't answer is To what extent you can take advantage of the very short duration of loading from each cycle? Can you allow for momentary yield a few tenths of a second without failure of the structure? I can't comment on the validity of Westergaard either, but I've never seen anything to replace it.
Regards,
DRG