I'm not sure I understand. When I say 3-D space, I mean, 3-D Euclidean Space or a different space with 3 variables.
3-D space in software is 3-D Euclidean Space. The 3-D space in particle emitters is probably also 3-D space, if you think of the coordinate system of the particles as consisting of an (x,y,z) or (r,z,phi) position.
If you mean if you have a particular simulator and the particles are bouncing / moving around, and they are doing so based on some initial conditions, and you want to compare how slightly different initial conditions will give different simulation results, then yes.
Yes, That is what I understood from your article. Cool Stuff!!!