On the validation of mixed-mode I/II crack growth theories for anisotropic rocks