Stay Cartesian, or go natural?: A comment on the article “Supernatural QUAD4: A template formulation” by C. A. Felippa [Comput. Methods Appl. Mech. Engrg., 195 (2006) 5316–5342]