Okay, let's look at the basic condition, eh?
Lift (and "induced" drag, which results from producing lift) calculate from velocity. The velocity term is squared. If there is a 10% increase in velocity, both lift and drag increase (1.1 squared = 1.21) over 20% compared to what they were at the original velocity. With larger changes, the effect is even greater, of course: 25% increased V causes over a 50% increase in lift and Induced drag (1.25 squared is 1.5625).
Velocity of a CL model depends on the flight radius and laptime. If the flier's chest is the center of the model's flight circle, we have about 2' from there to the handle, we have the lines length, eye to eye, and the length of the leadouts before we get to the spanwise CG, more or less. Say we have 60' lines I-to-I, and a 2' half span for leadouts. That's arm + lines + leadouts of 64'.
The outboard tip is another ~2' out, at 66' from the center. The inboard tip is, well, duhh, 62' from the center. 66' is 106.5% of 62'. So lift and drag at the outboard tip are 13.5% more than at the inboard tip. But there is also a steady variation of velocity across the entire span - it isn't just at the tips. Somewhere between the tips is a point where lift and drag are 'balanced' each side of the point... ...Where drag doesn't cause yaw, and where lift doesn't cause roll...
A friend who speaks calculus convinced me that the way to find that "dynamic mid-point" is to cube the inboard radius, cube the outboard radius, average them, and find the cube root of the average. Oooh, scary, right? Nah...
62 ^3 = 238328
66 ^3 = 287496
Average = 262912
Cube root of Average = 64.0625
.0625 feet = 0.75"
Since the model in this example has a 4' span, the structural midpoint would have been at 64' radius. It calculates 3/4" further outboard. Now, remember, this is how far the midpoint shifts; that means the inboard panel is ~1.50" longer than the outboard at the point where drag and lift are balanced inboard and outboard...
Other things are involved, of course, but this is the basic idea behind different panel lengths.