> Well, isn't that by definition true for any system of measurements? I mean, as long as we're talking about a simple straight-line distance in some real-dimensional space, it's going to be possible to measure that distance in meters (or yards, or whatever). Yes I know about fractal lengths, coastlines and so on.

The US yard is exactly 0.9144m such a short decimal expansion is highly unlikely when selecting two random units of measurements. It is short because we defined the modern yard in terms of meters, selecting a short decimal expansion that was still "close enough" to the old definition to allow tooling to remain the same.