No, this isn't that impressive. That's the point. Programming in Excel is not that difficult.
The integration is just Euler integration. Collisions are done with precomputed normals (the normals are computed in Matlab and imported as data into Sheets 3 and 4). The collisions aren't exact which is why at higher timesteps the particles noticeably interpenetrate the terrain.
The actual Excel macros only handle the drawing and copying the new (x,y, dx, dy) as the starting state for the calculation.
Loading more stuff…
Hmm…it looks like things are taking a while to load. Try again?