Some data:
At top speed, over 90% of power needed to keep the vehicle moving is used to overcome aerodynamic drag.
Power needed to keep a beetle body moving at constant speed on level wind-free road:
At 62 mph/ 100 kmh: 24 bhp ( 53 beetle 1130cc / 5,8:1 comp "30 sae hp"/24bhp/19kw)
At 72 mph/ 116 kmh: 34 bhp ( 63 beetle 1192cc / 7.0:1 comp "40 sae hp"/34bhp)
At 81 mph/ 130 kmh: + - 48/ 50 bhp ( beetle 1600 single / dual port)
Above data from various magazines that did speedo corrected tests using 5th wheel calibrated instruments.
VW's are geared so that top speed roughly (but not always), occurs near max power in 4th gear.
Karmann Ghia body did 5mph faster (77mph/ 116kmh) with same 1192cc 34bhp engine as beetle.
(Tested by same magazine on same level road, average of speed both directions). Ghia 1192cc had same gearing as beetle, maybe slightly taller gearing (to take advantage of it's aerodynamics), would make it 1 or 2 mph faster).
In fact, if so it would then have around the same top speed as a 1500 beetle, while only using 34bhp in stead of the beetle's 44bhp, (but obviously take much longer to reach that speed).
Type 3 body:
Intersting comparison is early squareback with original 1500 single carb (32PHN sidedraft 23.5mm venturi) 45bhp.
Same gearing as 44hp 1500 beetle. Tested by same magazine, same equipment, same place.
Accelerating to 50mph, the beetle was faster at 13.3 seconds versus the heavier squareback's 14.9 seconds. Beetle reached 70mph in 37.5 seconds. By that time, the type 3, even in squareback form, have already made up lost time and passed it 2 seconds earlier at 35.4 seconds.
Some calculated that notch and fastback had less drag than squareback, but, strangely, i could not find any data that supported a major difference. Think that a small flat spoiler (starting below the fastback's rear window,continuing level. Untill the rearmost part of car while slowly tapering inwards from the sides. Stopping abruptly there, Kamm effect). Think something similar at the rear would improve the Karmann Ghia as well.
What hurts the beetle and many other cars, is the rounded rear, which causes flow separation/turbulence. Compare a round sphere with another sphere cut in half. Facing the wind with it's rounded side, the "cut off" sphere (sharp edges facing rearwards) has much reduced drag compared to the rounded one. This is called the "Kamm effect".
Power needed to overcome aerodynamic drag rises exponentially as speed increases, so for a small increase in speed, a large increase in power is needed.
A bay or split bus roughly needs 50% more power to keep it moving at same speed as compared to a beetle.
Coefficient of drag:
Various figures related to aircooled VW's were quoted over the years. Calibration differences occurred between various wind tunnels and for the same wind tunnel over different years. Especially around the 60's and earlier.
Beetle: cd 0.46 to 0.48 mostly quoted, sometimes up to 0.49, and around the 1930's they quoted cd 0.39 (which was probably too optimistic at the time).
My guess probably 0.48 for 1958- onwards flat screen beetles. Maybe 0.46 for earlier versions which had the much smaller windscreen and more gentle curve at critical area near the top of windscreen.
Super beetle/1303 had cd 0.44 mentioned.
Baywindow and split bus- around cd 0.42 Obviously it has about 50% more frontal area, so total drag is more. (Westfalia bay was worse.) Around 0.39 was once mentioned for early split. (Probaly the pre-1954 barndoor version without the "peak" over the windshield). Vanagon: cd 0.45, and cd 0.44 with front spoiler. Vanagon is wider, so frontal area is more than bay.
Interesting study from the 1930's was the "Pillbug" by professor Schlor. Built before the beetle chassis was available, it used the chassis from a Mercedes 170h rear- engined car. It was a 7-seater and cd was somewhere around 0.15.
Obviously, below 30mph aerodynamics is of little importance. Then reducing weight (acceleration/deceleration) in city driving conditions is important