it's faster because it's less data to send down the bus.
consider 4,000 quads. = 4,000 4,000 vertex definitions. the rest filled in on the card or the gpu as bot mentions above. compared to 4,000 x 6...(Wait..i was never any good at maths...)
consider 4,000 quads. = 4,000 4,000 vertex definitions. the rest filled in on the card or the gpu as bot mentions above. compared to 4,000 x 6...(Wait..i was never any good at maths...)