We performed comparisons of our interstellar extinction models, which are based on the hypothesis that gas and dust are homogeneously mixed and make use of the dust-to-gas ratio to make extinction predictions, with samples of objects for which other methods of estimating extinction are available in the literature: giant elliptical (gE) galaxies, star clusters, and the extinction toward the Galactic center. We used the same sample of gE galaxies that Burstein selected in 2003 for a comparison he made between the extinction predicted by Burstein & Heiles and by Schlegel and coworkers, to extend the comparison to our models. Due to the quality of the sample of gE galaxies, we slightly tuned the parameters of our model A, getting what we call the improved model A2. We found 3 σ-limited rms differences in E(B - V) around 0.017 mag in the comparison of the predictions of this model with the two methods mentioned above. Our comparisons take into account the zero points recommended by Burstein. The correlation coefficient obtained in the comparisons is around 0.86. These results bring to light that our model can be safely used for extragalactic work in order to estimate the extinction in our Galaxy, as an alternative method to the predictions of Burstein & Heiles and Schlegel and coworkers. In the comparison of our results with the globular cluster data we found rms differences in E(B - V) equal to 0.32 and 0.30 for our models A and S, respectively. For the open clusters we made comparisons using different samples; the rms differences were typically 0.25. In the Galactic center, comparing our model A with the extinction map in the K band provided by Dutra and coworkers, we found an rms difference in A_K equal to 0.16, which is relatively small, considering the high extinction values in that region.