Detailed statistical rate calculations combined with electron capture theory and kinetic modeling for the electron attachment to SF(6) and detachment from SF(6)(-) [Troe et al., J. Chem. Phys. 127, 244303 (2007)] are used to test thermionic electron emission models. A new method to calculate the specific detachment rate constants k(det)(E) and the electron energy distributions f(E,epsilon) as functions of the total energy E of the anion and the energy epsilon of the emitted electrons is presented, which is computationally simple but neglects fine structures in the detailed k(det)(E). Reduced electron energy distributions f(E,epsilon/<epsilon>) were found to be of the form (epsilon/<epsilon>)(n) exp(-epsilon/<epsilon>) with n approximately = 0.15, whose shape corresponds to thermal distributions only to a limited extent. In contrast, the average energies <epsilon(E)> can be roughly estimated within thermionic emission and finite heat bath concepts. An effective temperature T(d)(E) is determined from the relation E - EA = <E(SF(6))(T(d))> + kT(d), where <E(SF(6))(T(d))> denotes the thermal internal energy of the detachment product SF(6) at the temperature T(d) and EA is the electron affinity of SF(6). The average electron energy is then approximately given by <epsilon(E)> = kT(d)(E), but dynamical details of the process are not accounted for by this approach. Simplified representations of k(det)(E) in terms of T(d)(E) from the literature are shown to lead to only semiquantitative agreement with the equally simple but more accurate calculations presented here. An effective "isokinetic" electron emission temperature T(e)(E) does not appear to be useful for the electron detachment system considered because it neither provides advantages over a representation of k(det)(E) as a function of T(d)(E), nor are recommended relations between T(e)(E) and T(d)(E) of sufficient accuracy.