It does seem a poorly worded regulation - perhaps it is flaunted because it is open to interpretation.
As John says, the surface (land) reflecting bright moonlight is more than sufficient to maintain attitude and orientation whereas the same lighting over a flat calm sea might not.
A single point light source on the land (especially on a dark night) is not enough to maintain attitude and orientation (autokinesis) but it would seem to meet the stipulated criteria.
A combination of a visual horizon and sufficient visible features on the ground is what is really required for safe night VFR flight - so why doesn't the rule say so?