The effect of the positive $Q$-value neutron transfer (PQNT) channels on the subbarrier fusion process is still far from a clear understanding and hence is worth further investigation. To this end, we have measured the fusion excitation function for $^{32}\mathrm{S}+^{94}\mathrm{Zr}$ at energies near and below the Coulomb barrier with rather good accuracy, and deduced the experimental fusion barrier distribution. The reaction system $^{32}\mathrm{S}+^{94}\mathrm{Zr}$ is a good candidate for the evaluation of PQNT effect because it has multineutron transfer channels with positive ground-state ${Q}_{xn}$ values as well as the relatively weak low-lying ${3}^{\ensuremath{-}}$ vibrational state in $^{94}\mathrm{Zr}$. Our results show that the fusion cross sections of $^{32}\mathrm{S}+^{94}\mathrm{Zr}$ are strongly enhanced at the lower-energy region. By means of a cross comparison of the $^{32}\mathrm{S}$,$^{40}\mathrm{Ca}+^{90,94,96}\mathrm{Zr}$ systems, further evidence is given that the positive ${Q}_{xn}$-value neutron transfer channels, rather than the strong low-lying ${3}^{\ensuremath{-}}$ vibrational state, bring about the additional subbarrier fusion enhancement. Relative to the coupled-channels results with couplings to the low-lying inelastic collective excitations, an unexpectedly large enhancement for the subbarrier fusion cross sections is observed in $^{32}\mathrm{S}+^{94}\mathrm{Zr}$ compared to $^{32}\mathrm{S}+^{96}\mathrm{Zr}$. Some existing theoretical models have been discussed, but a good explanation cannot be achieved. Exact coupled-channels theory and more experiments on transfer reactions themselves are strongly desired to comprehensively understand the effect of transfer on fusion.