One thing I've learned is that small changes can make better planes.  For combat though, the first things are: a reliable engine that starts in one flip, holds a needle valve setting, somewhat affordable, and durable.  A good pit crew.  Depending on the event, the pit crew can be vital.  In AMA Fast combat the pit crews probably win at least as many matches as the pilots by getting up first. Being able to get to, retrieve, clean up, fuel if needed, and change the prop in a minute or less.  The plane isn't quite as important, but a good design in good trim will help a lot.

It's kind of a shame, but with F2D Fast the plane has lost a lot of innovation.  In fact, designs have been optimized to such a degree that it's basically a one design event.  Most flyers buy their planes ready to fly and there are only two basic designs- one with the stab hanging behind the wing on a small hinge, and one with a flipper elevator sewn to the back of the wing.  The integral flipper seems to offer an advantage in smoother control while the stabilator version reacts faster with less handle movement.

But even today, small changes in the plane can make a difference.  I've found through multiple successful design that roughly a 5% change in span, wing area, aspect ratio, or engine power can make a useful difference in performance.  Take two planes with similar construction- extending a 36in. span to 38-39in. with a narrower chord will make the plane faster in maneuvers.  Knocking an ounce off of any design will let it turn just enough tighter to give an advantage. The Gotcha 500 design I used for several years was a far out compromise with and advantage in turn radius(+10-25% wing area), wing span(10-20% span), and an engine that could turn any reasonable prop fast enough to go 110-115mph, which was all I could handle.  The down side was that it was so lightly built it would barely last 20 flights before starting to get too flexible from fatigue.

The Nelson engine opened up new possibilities.  I built a 550 sq.in., 56in. span design that flew very well.  It was especially good against the ARF fast planes.  They were restricted to about 46in. span and needed quite a wide chord.  That design had plenty of speed, but tight turns slowed it down a lot due to the low aspect ratio.

I've always found a higher aspect ratio is faster than a low one. Front drag on a good airfoil will always be less drag than a wide chord.

I'm no Obi-Wan.  I've just made a lot of mistakes.  If you want real results maybe say a few prayers.

Direct answers are that every square inch slows you down.  Unless you go with hyper thin airfoils- half an inch thick or extremely thick ones, say 25% the drag from airfoil shape is fairly limited.  That's what airfoil design is about.  For combat right now the formulae have been pretty well established.,
The pros classify drag into categories- some are more important than others.
Combat planes have relatively fixed power so anything to reduce drag helps.
Parasitic drag-  a fancy term for drag that comes just from the plane moving though air.
       Skin Friction drag- just from the air moving over the surface.  Reynolds number comes into play   
       Interference drag- from all the things sticking out of the wing and tail- rivets, struts, joints, fuselages, etc
       Form drag or pressure drag- drag from the shape of the plane. All the other bits have been pretty constant since the early '60's- engine and nacelle, boom or booms, tail surfaces and controls.
Profile drag- from the shape of the airfoil, primarily.  A racer has a very thin wing to go faster, a combat plane has an intermediate thickness.  Heavy lifters may have fairly thick wings for strength.

Lift induced drag- what happens when you move the elevator.  The angle between the wind and the airfoil(front of leading edge to the trailing edge) is angle of attack or AOA.  Somewhat mysteriously, as you increase the AOA lift and drag go up in a straight line fashion up until about 15° when wings stall.  The airfoil affects primarily how the wing stalls.  A sharp leading edge will cause an abrupt loss of lift and make the plane do nasty things.  A blunter leading edge(anything over about 1/4in. radius)  will do things like drop a wing tip because the wing isn't perfectly built and one part stalls before the rest.
Since about the mid to late 70's the most popular airfoil shape  is an ice cream cone- a very blunt leading edge back to spars set 20-25% of chord and then a slight curve to a straight line back to the trailing edge.   This shape will generally softly stall one tip or the other making the plane flop a bit but if you don't force it the pilot can recover easily.  It just means the pilot has too much control.

My little experiment comes under Form drag- the shape of the wing.  Increasing the chord added 60 sq.in.- 12% .  The aspect ratio went from 4.5 to 4.11, an 11% reduction.  In level flight the wing area adds a fair amount of drag.  Drag is minimum(in our case of fixed power) when the plane hits top speed in level flight.  That 12% takes a lot of power- about 25% more to get the same speed as the smaller plane.  Drag goes up as the square of the speed or conversely more drag slows the plane down as the square of the speed. Power functions are hard to design around.

If I'd adjusted everything to get a higher aspect ratio with the same increased wing area it would have been a competitive flyer- a bit slower on the same engine but faster with in the turns.  Once the Nelson 36 came out you could make almost anything fly faster than you could fly.  I top out at about 110mph.

The aspect ratio takes charge in maneuvers.  The air spills around the wingtips when the plane is turning.  Cdi = (Cl^2) / (pi * AR * e)  I know everybody hates math the the little equation says the when aspect ratio(AR) gets larger the induced drag- Cdi - gets smaller.  As a practical matter
AR = s^2 / A   span squared/wing area  limits us to a range of 4.5-5.5.  F2D planes are mostly right around 5.   Keeping the same wing area you could build a 420 square inch plane with a 48in. span and cut the drag increase in a turn about 5%.

Michael Hoffelt designed his Monoboom with an aspect ratio of 7.  Long span for it's weight.  Everybody who flew it said it flew great, especially in maneuvers.  On the minus side the 3 different sizes- Half A,  F2D,  and AMA Fast were pretty fragile.  It needed a silk or silkspan covering for stiffness.  It made a great demonstration plane for fast and furious turning though.
At the other extreme was the Sassy Saucer, or Walt Williamson's Flying Flounder, which had about a 16in. span and maybe 18in. long.  Great for demonstrations but not really cut out for combat.

I've always found a higher aspect ratio is faster than a low one. Front drag on a good airfoil will always be less drag than a wide chord.

I agree.

Andre, I believe(somebody else may Know) that without knowing at least the focal length used to take the picture you can't estimate the parallax effects from the camera film not being exactly parallel to the plane image.

One thing you could try is to import the photo into a 3D CAD drawing and measure the difference in length of the two tip ribs and the rib underneath the beer can.  Beer cans have stayed pretty much the same height for a long time.  Then you can draw the shape of the wing as seen.  Rotate it until you find the correct angles and distance from the veiwpoint in the photo to make the wing a rectangle.  There's also a yardstick just behing the plane.  You can see how closely the markings match up with a 1in. line.

The airfoil is probably just a french curve style mated to the leading edge shape.  I doubt your friend was working from NACA records.  They weren't readily available for years after the war.