A recent article [Surface Sci. 209 (1989) 379] has attempted to improve the agreement between our original X-ray diffraction data [Phys. Rev. Letters 54 (1985) 1275; Surface Sci. 186 (1987) 499] and the proposed distorted vacancy model by the introduction of an additional, partially occupied atom in the unit cell. Here we show that almost as much improvement can be obtained by introducing second-layer displacements into the original structure. This raises questions of uniqueness in crystallographic structure determination and the level of detail attainable without overinterpretation of data.