f mod 8 pics attached for odd x+n = 13 showing (hopefully) more consistent symmetry with n, d, and f values.
The f mod 8 = 0 has been updated to incorporate VA's idea of not adjusting the nn portion. >>5665
Additional images have been created for f mod 8 = 1, 2, 4, 5, and 6.
In most cases, it seemed to make sense to fill the center square with 1 taken from (f-1). See the yellow center square in images for 0, 2, 4 and 6.
In the cases of f mod 8 = 1 and 5, however, the symmetry looked better taking the 1 center square from nn. See the green center square in those images.
Based on these adjustments, a few of these mod cases start to look increasingly similar. 1 and 2 have some overlaps. As does 5 and 6.