We investigate ultrascaled $\langle 100\rangle $ -, $\langle 110\rangle $ -, and $\langle 111\rangle $ -oriented p-type silicon nanowire transistors using a quantum transport simulator based on the: 1) six-band $\text{k}\cdot \text{p}$ method and 2) sp $^{3}\text{d}^{5}\text{s}^{*}$ tight-binding model. The hole transmission probability from source to drain at low gate voltages shows discrepancies between both models in the $\langle 110\rangle $ and $\langle 111\rangle $ cases. The origin of this phenomenon can be traced back to the nonparabolic band behavior of the imaginary dispersion that is not captured by the six-band $\text{k}\cdot \text{p}$ method. In order to accurately reproduce the full-band (FB) wave function attenuation in the bandgap, the hole energy must be corrected. This approach is validated by comparison with FB calculations. The results suggest that the observed failure of the six-band $\text{k}\cdot \text{p}$ method at ultrashort gate lengths can be avoided, thus extending the applicability of this computationally efficient model.