I should have posted this some time ago.  I built a larger 5×5 grid using the same Neodymium magnets that I used in an earlier experiment.  The flux is a lot less strong than when they were close coupled.  This tallies with conversations that I have had with electromagnet engineers around the nature of the grid, and how it is linear at any given point.  The most exciting result from this experiment is that the fluxs here were similar to what I was achieving with my shorted lived coil experiment.  I used a square format here, but I have since become more interested in a hexidecimal grid system.