Yeah, I read some of that - messy! Ingenious as it is, the Rip-o-meter is a fairly crude way to measure a switch - adding weights a fraction of a gram at a time isn't going to improve its accuracy much! Friction probably amounts to a few grams even on a linear switch (and I guess goes part of the way to explaining the gap between the down- and up-stroke lines on the graphs of black and red).
Here's a graph of my findings so far from measuring just the springs, by compressing them to various lengths on some scales:
Clearly they're falling into three groups. The 'medium' ones are all roughly the same force at 8mm, which I think is about the actuation point. However, it looks like the spring rate goes up as you go from green, through black, to clear. That difference in slope might be easier for a sensitive finger to detect than the absolute force at any point.
Wish I had a red spring or two to test, but I'm in little doubt that it would match the brown and blue almost exactly (i.e. within measurement error).
