[en] This paper discusses two common approximations used to calculate average fission cross sections over the compound energy range: the disregard of the W_I_I factor and the Porter-Thomas hypothesis made on the double barrier fission width distribution. By reference to a Monte Carlo-type calculation of formal R-matrix fission widths, this work estimates an overall error ranging from 12% to 20% on the fission cross section in the case of the "2"3"9Pu fissile isotope in the energy domain from 1 to 100 keV with very significant impact on the competing capture cross section. This work is part of a recent and very comprehensive formal R-matrix study over the Pu isotope series and is able to give some hints for significant accuracy improvements in the treatment of the fission channel. (authors)