I have an instron at work, there's a small chance that I could rig up a switch tester in order to correlate the force profile during depression and the point of actuation. I'm not sure if that information is already available though, seems like it should be from Cherry.
Also do you only want to collect data on certain switches and find an average for hookes modulus? Do you see a benefit in swapping the springs and stems rather than just using an assembled switch? If you expect enough variation in just the housings, I feel that the data may not represent a very large population. Lastly, what kind of statistical sampling plan did you have in mind?