Scampy, There are actually several reasons to do this.

While some newer transmitters have the ability to modify the servo response curve, older or simpler units require exactly this type of set-up. Suppose for example, rather than a linear response from the servo you require a logarithmic one? Or, rather than a smooth rotation, 10 discrete positions... with a microprocessor between the receiver and the servo virtually any response may be programmed to affect the servo - or servos. Really, nothing requires that the output be limited (or similar) to output to more than one servo. also, timing may be a reason - flip the stick to half... and the servo follows in a timed arc - "realistic" flying, where pilots try to mimic the look of real aircraft while flying is an aspect of this hobby many enjoy. I suspect that with proper timing and knowledge complex "3D" moves as well as many visual effects (lights, landing gear, bomb drops, camera activation) might be customized to the switches, sliders, and knobs on today's transmitters.

I have a pretty nice Futaba that limits the response of one servo based on the position of another - say, for example, to keep flaps from deploying at full speed or reducing control surface response at high throttle settings. Many of today's hi-tech models - from "Wings" to "V-tails" require complex control interaction to fly. And, of course there are any number of ground and water based models - and robots - that would only be possible with this type of control.