Disadvantages:

• Defining size with object axis length as a size reference does not assist reasoning about object interaction, because it obscures the relative sizes of shape features on different objects (since shape features on each object are related to the axis on that object only). Reference to axis length would be more useful in solving problems in CAD, object classification, or other areas where objects are considered one at a time.
• Description of the size of shape features by comparison to the major axis fails to recognise the common case where two shape features fall within the same subdivision of the axis size, but have a functionally important difference between their own sizes.
• Although object axes can provide functional information, local shape features on the object boundary are far more significant where close interaction between objects is occurring, and the relative orientation of features is therefore more important in this situation than axis orientation is.
• The methods for describing location and orientation of shape features are rather clumsy. They describe location and orientation sufficient that object shape can be largely reconstructed (by a human) from the description, but they are difficult to use when reasoning about motion is being carried out.