There are various different definitions for the triaxial deformation parameter "γ ". It is pointed out that the parameter conventionally used in the Nilsson (or Woods-Saxon) potential, γ (pot:Nils) [or γ (pot:WS)], is not appropriate for representing the triaxiality γ defined in terms of the intrinsic quadrupole moments. The difference between the two can be as large as a factor two in the case of the triaxial superdeformed bands recently observed in Hf and Lu nuclei, i.e., γ (pot:Nils) ≈ 20 • corresponds to γ ≈ 10 • . In our previous work, we studied the wobbling excitations in Lu nuclei using the microscopic framework of the cranked Nilsson mean-field and the random phase approximation. The most serious problem was that the calculated B(E2) value is about factor two too small. It is shown that the origin of this underestimate can mainly be attributed to the small triaxial deformation parameter γ ≈ 10 • that corresponds to γ (pot:Nils) ≈ 20 • . If the same triaxial deformation parameter is used as in the analysis of the particle-rotor model, γ ≈ 20 • , the calculated B(E2) gives correct magnitude of the experimental data.