Recently developed free-energy density functional theory (DFT)-based methodology for optical property calculations of warm dense matter has been applied for studying L-shell opacity of iron and chromium at T=182 eV. We use Mermin-Kohn-Sham density functional theory with a ground-state and a fully-temperature-dependent generalized gradient approximation exchange-correlation (XC) functionals. It is demonstrated that the role of XC at such a high-T is negligible due to the total free energy of interacting systems being dominated by the noninteracting free-energy term, in agreement with estimations for the homogeneous electron gas. Our DFT predictions are compared with the radiative emissivity and opacity of the dense plasma model, with the real-space Green's function method, and with experimental measurements. Good agreement is found between all three theoretical methods, and in the bound-continuum region for Cr when compared with the experiment, while the discrepancy between direct DFT calculations and the experiment for Fe remains essentially the same as for plasma-physics models.