So - You want to learn to fly?

That's how we used to do it, the problem is, without careful link arrangements you can't have aileron differential. Given the cost and strength of servos it makes more sense to split the load. Each on a separate channel, then you can have differential, flaperons and anything else you want to try, with the added bonus that should one servo fail you will still be able to control the aircraft.

Plus - the old torque rod method  unless everything is extremely well made, you can get 'slop' - microscopic slack in the various linkages, and if the wong (Chinese!?) gauge of wire is used, it can twist under load. Both of these will lead to control surface flutter, which in turn usually leads to something, usually the surface,  falling off = crash!
You can also find methods where a single central servo pushed or pulled two rods, one to each wing side, that then coupled to a 90' bellcrank, which then coupled to the control horn. That way there is no wire rod to twist, but introduces three more pivot points instead - the two connections to the bellcrank AND the mounting pivot. It also means you are limited to size of wing that will fit into your chosen transport method, as it is impractical to keep disconnecting the rods inside the wing! Another disadvantage is you need a more powerful therefore bigger servo, which when mounted in the wing, will stick out and be prone to crash and transport damage, if it doesn't knock a hole in the side of the fuselage when the wing twists!
Also it makes the comment in the second half of Klippy's comment impossible
With a servo in each wing, directly driving a straight rod to the control horn, there is less room for slop and flutter, and it also means that a physically smaller servo can be used - as Klippy put, you have halved the load on the servo.

Here's another observation, I imagine the builder modified an earlier installation, putting the servos so close to the fuselage. It would be better to put the servos halfway along the full width aileron, then any twisting moments (there will be some) are halved. Moving a long surface from one end is never good engineering practice.
We're not trying to dis you Magpie, it's just telling it like it is. You've embarked on a long journey, best you start with the right information.

with the servo on its side partially within the wing - -
That will give you a very weird assymetric differential - one rod pushing straight, and the other pushing up at an odd angle! And - unless the two torque rod horns are very close together, you are building in a mechanical twist, again.  ((like we did last summer!)) 

If you insist on a single servo and torque rods, the servo is best mounted  so the output disc/arms are parallel to the surface of the wing, with one rod from each side going to one torque horn. That way the loading is along the length of the servo body, and both rods are parallel to the wing surface.
You will need to make a V notch along the trailing edge of the wing AND in the aileron, to accommodate the torque rods, or you'll have a big unsightly gap.
Make sure you degrease and rough up the outsides of the tubes, or the epoxy won't bond to them, but with a DROP of oil on the wire, so any stray epoxy won't glue it all solid!
It would be mechanically much simpler with one servo on each wing panel, driving one aileron each via a short rod, Plug the two servos into a Y lead, and the Y lead into the servo output of the rx, using 'keepers' so they can't accidentally (or through vibration) come unplugged.
Then, your two short rods connect to the servos - each on the (say) outside arm, which will automatically give you ailerons going in opposite directions! (You can use both rods to the inside arms if you wish, same effect, but opposite surface deflection!).
But it is your choice - it is your model!

Magpie, I think you will find that 20' each way is more than enough, unless you want to fly aerobatics!!
There are several factors involved, though - area of the control surface vs area of the wing, chord of the surface vs chord of the wing, and airspeed - being the main ones.
Should you get the opportunity to look at a fast turbine model, you will find that the control surface movements are a very few degrees, whilst those of a very slow indoor 3D one look ridiculously huge.
Think of your control surfaces as brake pedals - every deflection from neutral is creating increased drag - which slows the aircraft down.

My preference is to have a lot of exponential dialled in, so that movements around neutral are small, but at full stick deflection, I get a lot of control deflection, thus the model will potter about like a trainer, but at the same time can perform all kinds of gyrations, without the bother of rate switches (and trying to recall which switch I set the rates on, when something goes wild!) and which way was on or off!

However some radios have voice functions that are programable and tell you what you have selected, even down to an adjustable delay in operation should you flick the wrong one.

I'm trying to work out what you mean -
- - - - should I position the aileron next to the fuselage - - - - -
Are you talking of the aileron - or the aileron horn?
The aileron itself should be out near the tip, for best effectiveness, with the horn that is driving it roughly central to the length (the span) of the surface, to mnimise the twisting effect of the air pressure when in flight.

Now, all those numbers on your sketch - what do they signify?
I'm assuming you are talking metric, so if your root chord is 4 (cm?) - just over 1.5 inches - that is a very tiny wing!
Then what are the 32 and 9.5? (16.1/4 inches total) That's a pretty small half wing!
What is the 8 for?
And if the root chord is 4, then the tip must be less than 1 (aout 3/8ths of an inch) if that sketch is anything like scale
What are the three parallel lines? and what are the two concentric circles?
I fear that you are overcomplicating and overthinking everything!

A typical aileron is about 50% of one half wingspan and about a quarter of the chord, positioned slightly in from the wingtip, with the control horn in the centre of the surface - but with the holes for the rod positioned vertically over the hinge-line - as shown in Q's post on the 27th