That's very cool! Thanks for posting it. Are you a professional modeler of some sort?
I would just caution that the apparent spacing in 2D is closer than the reality of 3D, if you just took the ratio of water and ions and mixed them together in 2D (not sure if that's what you did or not). Extrapolating to 1D (just spacing all the water and ions on a line), the relative spacing error would much worse (about 1 ion per 55 water molecules in a line).
Thanks! Yup, computational biophysicist here!
The simulation was in 3D, but I agree that it looks closer when projected in 2D. They also look closer when the solvent isn't shown as spheres (but they would block out all the ions). Here's an updated video rotating the system, with the ions only. I should note that I updated the van der Waals radii to reflect the radius of the charged state (i.e. instead of Na, radius is Na+)
Showing everything as spheres, including water (orange and white now), the ions look more sparse.
And just because I'm having too much fun, I calculated the distance between each ion and its closest ionic neighbor for the last 10 frames of the sim (results are approx since its a short sim in a small box). As others have pointed out, ions are not uniformly distributed! I've also broken this down on a per-ion basis. For comparison, the dodecahedron box is ~6 nm in width.