Could that be derived somehow, or would you need a measurement tool to handle it directly?
It's my chief frustration with Lethal Squirrel's Cherry MX animations: they don't depict the way your finger jumps as you go past actuation.
I finally created part of (just the key, without the accompanying graph, and downstroke only) an animation to show this effect, but it came out surprisingly ineffective:
[wiki]Introduction to keyboards[/wiki] (towards the bottom)
It goes back to my ADSR model of force curves, and whether you have path return or path divert. My observation is that "path divert" switches feel inherently linear, including MBS and Futaba clicky. You can see the Futaba clicky switch force graph at the bottom of this page, and it's a path divert design:
http://www5f.biglobe.ne.jp/~silencium/k ... 8pc88.html
However, there's another distinction: positive-only (tactile peak) and positive-negative (tactile peak followed by a tactile trough).
The difficulty is that you need two things to be able to make these assessments:
a) Force curves of every known switch type
b) A huge collection of keyboards, as loose switches are worthless for investigating feel
I don't have space for the keyboards or any hope of measuring them, so I won't ever be able to progress these ideas in terms of feel, and classification based solely on visual inspection of the force curves isn't sufficient.