Virtual Deep Seismic Sounding (VDSS) has emerged as a novel method to image the crust–mantle boundary (CMB) and potentially other lithospheric boundaries. In Part 1, we showed that the arrival time and waveform of post-critical SsPmp, the post-critical reflection phase at the CMB used in VDSS, is sensitive to several different attributes of the crust and upper mantle. Here, we synthesize our methodology of deriving Moho depth, average crustal V_p and uppermost-mantle V_p from single-station observations of post-critical SsPmp under a 1-D assumption. We first verify our method with synthetics and then substantiate it with a case study using the Yellowknife and POLARIS arrays in the Slave Craton, Canada. We show good agreement of crustal and upper-mantle properties derived with VDSS with those given by previous active-source experiments and our own P receiver functions (PRF) in our study area. Finally, we propose a PRF-VDSS joint analysis method to constrain average crustal V_p/V_s ratio and composition. Our PRF-VDSS joint analysis shows that the southwest Slave Craton has an intermediate crustal composition, most consistent with a Mesoarchean age.