That sounds like a little too much added complexity than a simple spring to me. If anything if something like that works well, it could be something we could retrofit later... (yes, I'm just a little impatient
)
If I am visualizing the scenario right...
The bottom shoulder buttons have the luxury of a deep well to sit over/in where a cylinder spring can simply push the plastic bit back outward. Behind the button is a convenient cavity in the plastic that can be used for this. Bonus. If a peg is created for the back side of the shoulder button itself, a coil spring can ride on this. Solved.
The top shoulder buttons, though, only have the relatively weak spring-back from the switch itself. Behind them is the top side of the mainboard - there isn't anything nearby to seat a spring against - no cavity, just board. I wish I had a set of these plastics or a rotatable model of the upper shoulder button - that would help me eliminate bad ideas as they pop up. But as I don't have those, a few of those likely bad ideas...
1. Motherboard mounted spring guide. Next to the top shoulder button switch is there room to mount a 'horn' to the board. Essentially a 90* hook of just large enough to be sturdy wire with one end pointed toward the switch activator pad to act as the spring guide. Any nearby through-hole ground points?
2. Silicone booger. There is a defined cavity/space/shape to where that upper switch lives/exists. Spray some cooking oil on the parts (for release) then fill the cavity with silicone. Let it set. Pull it apart then trim op the booger so that it fits the space properly - and is a mold-friendly taper. Test. Should be a bit squishy yet reflex/spring back. If it is too firm, cut a bit out of the sides so it squishes easier.
3. Levered flat spring. The cavity where the bottom shoulder springs will get located may be able to be the 'firm place to stand' to move/push on the upper spring by using flat thin spring steel. The easiest place to get small pieces of this stuff is to -very carefully- disassemble a badge retractor. The coil of flat steel inside is VERY sharp and seems to move of it's own accord in a finger-slicing eye-blinding and pain inducing spiral of death. Wear gloves & goggles if you take one apart. But - if that material is anchored in the cavity - it might be thin enough to run through between the board & the switch to apply back-pressure on the switch.
All likely bad ideas, but maybe they could help find a good idea?