In Li et al. [phys. Rev. A 98, 020502(R) (2018)] it was claimed that the model-potential computations of the Lamb shift on the P-2(1/2) - P-2(3/2) fine structure in fluorinelike uranium lead to a discrepancy between theory and experiment. Later, it was reported by Volotka et al. [Phys. Rev. A 100, 010502(R) (2019)] that ab initio QED calculation, including the first-order one-electron QED contributions and the related effects of two-electron screening, yields the result which restores the agreement between theory and experiment and strongly disagrees with the model-potential Lamb-shift values. In the present paper, the model Lamb-shift operator [Shabaev et al., Phys. Rev. A 88, 012513 (2013)] is used to evaluate the QED effects on the P-2(1/2) - P-2(3/2) fine structure in F-like ions. The calculations are performed by incorporating this operator into the Dirac-Coulomb-Breit equation employing different methods. It is demonstrated that the methods, based on including the Lamb-shift operator either into the Dirac-Fock equations or into the calculations by perturbation theory, lead to the theoretical results which are in good agreement with each other and with experiment. The restriction of these results to the first order in the QED effects leads to a value which agrees with the aforementioned ab initio QED result.