[MaciekA]

I have actually experimentally tested the backwards-forwards non-linear spring theory and have found there to be no appreciable difference in compression from reversing a non-linear spring.

I've heard this opinion echoed elsewhere, along with the theory that most people who have noticed this "effect" in actuality fixed some other compression issue while flipping the polarity of the spring.

I'm willing to accept experimental evidence that contradicts my findings if there's some other explanation (relating to perhaps porting or AoE or something), but for the moment, the reversed non-linear spring theory holds absolutely no water for me.