A different explanation

At Marginal Revolution, they pass on this study which shows that we overestimate the volume of a tall, thin glass and underestimate the volume of a short, wide glass. The explanation -- we tend to think of size in terms of vertical rather than horizontal measurement.

That may be, but I have a different hypothesis. We tend to be confused by greater surface area. I don't think it is a vertical/horizontal issue. A sphere or a cube will appear smaller than objects which are compressed in one of the dimensions. Take a ball of clay and roll it out. The flatter shape will appear larger than the ball (although it obviously has the same volume) because the surface area will be far greater.