The analysis of S{I} recoupling experiments applied to amorphous solids yields a heteronuclear second moment M(2)(S-I) that represents the effective through-space dipolar interaction between the detected S spins and the neighboring I-spin species. We show that both M(2)(S-I) and M(2)(I-S) values are readily accessible from a sole S{I} or I{S} experiment, which may involve either S or I detection, and is naturally selected as the most favorable option under the given experimental conditions. For the common case where I has half-integer spin, an I{S} REDOR implementation is preferred to the S{I} REAPDOR counterpart. We verify the procedure by (23)Na{(31)P} REDOR and (31)P{(23)Na} REAPDOR NMR applied to Na(2)O-CaO-SiO(2)-P(2)O(5) glasses and biomimetic hydroxyapatite, where the M(2)(P-Na) values directly determined by REAPDOR agree very well with those derived from the corresponding M(2)(Na-P) results measured by REDOR. Moreover, we show that dipolar second moments are readily extracted from the REAPDOR NMR protocol by a straightforward numerical fitting of the initial dephasing data, in direct analogy with the well-established procedure to determine M(2)(S-I) values from REDOR NMR experiments applied to amorphous materials; this avoids the problems with time-consuming numerically exact simulations whose accuracy is limited for describing the dynamics of a priori unknown multi-spin systems in disordered structures.