Conjecture on the affine full-dimensionality of D-optimal factorial designs

Let $r$ denote the number of runs, and distinguish affinely full-dimensional factorial designs from subset fractional factorial designs. The D-optimal design conjecture. The $D$-optimal design is affinely full-dimensional factorial when

$$r\equiv 5,6,7\pmod{8},$$

and it is subset fractional factorial when

$$r\equiv 0,1,2,3,4\pmod{8}.$$

The conjecture records the periodicity observed in the table of known maximal determinants and predicts the type of $D$-optimal design according to $r$ modulo $8$. The source gives no resolution or status evidence.