Magnets arrived (rectangular), linear hall sensors (49E) as well and I found some springs. Some tests to get a feeling about sensitivity and distances. 

Might work so I started to sketch a demonstrator enclosure. I am still not sure how to place the magnets and halls exactly. Either I measure 6 times the "angled" z distance to the top or 3 times the z displacement and 3 times the in plane displacment like the optical system by connexions space mouse.

Not really sure, so I just stared with a first guess and printed the parts to get some more experience.

I had 20mm long springs with inner diameter of 5mm. The soft ones had a wire diameter of 0.4mm and harder ones of 0.8mm. I just sticked the springs to the poles without glueing

First impressions:

• Magnet fixture needs more distance towards springs and are to thin (break). A mechanical fixture of the hall sensors will be helpfull
• Soft springs seamed too soft, hard ones somehow OK. 
• For the hard springs, I think it will be hard to lift the top w/o lifting the base plate. Difficult to test without a heavy base and fixed springs. Maybe preloded?
• Stoppers needed, but I will do them later, initial test work well without