Thank you for linking the comparison to S2! The big difference to me seems to be the shape of the cells: hexagonal for H3, square for S2. H3 hexagonal is bett

Thank you for linking the comparison to S2! The big difference to me seems to be the shape of the cells: hexagonal for H3, square for S2. H3 hexagonal is better for seeing more neighbouring cells. S2 is better for being referenced with a numeric coordinate system, and uses a fractal space-filling curve to generate subdivisions. https://s2geometry.io/devguide/s2cell_hierarchy https://s2geometry.io/devguide/s2cell_hierarchy Fractals are super efficient (a binary tree is a fractal used for search) so I feel like I prefer S2 in this case. How are fractals being used in H3, and could other cell geometry be more efficient than the space-filling curve fractal used in S2?