Not when dealing with non-ideal springs that are submerged in fluids that act as dampers (the water surface or the rig flapping about). In such case the flow of leaked energy (grag) can be significantly greater.

In our case it is not just the spring itself (the beams etc) that we are concerned with. In our case these are attached to large surfaces that flap about in the water or air. Fit a spring to an object with a large surface area (and low weigt) and see whether you can still recover 99% of the stored energy while having it oscillate.

Other then that do this mind experiment.

Lets lay our boat motionless in the water. Then flex the platform, forcing one bow 40 mm down. Let go of the depressed bow and see whether the boats starts to move forward. I say that I see waves being created and the boat and rig shaking about but not much forward motion if indeed any. Now use the same amount of force (and travel) needed to flex the platform to push the boat forward. Spot the difference in the resulting motion ?

So our boats are not designs that recover much of the energy stored in flexing the platform at all.

Wouter