NB, I haven't figured out how to use the matching feature that comes with Eznec, but IMO and within reason, I don't think correcting the SWR to show a better match makes much difference in the performance of a model, at least not so one can easily tell just using our radio.
Furthermore, I see virtually nobody using matching with their models in support of their antenna comments in professional type antenna literature, including W8JI and Cebik. This all may say something about why we don't hear much discussion on matching issues talked-about in the antenna modeling literature also.
I have made an I-10K model where I physically added the trombone matching device, and it does show to positively effect the pattern and and improves the match at the feed point. The addition of the trombone matching device also shows to add some horizontal radiation to the pattern. So, if this model is close to correct...then this big I-10K trombone tuner may be what accounts for the improved performance many CB operators report over other 5/8 wave type 11 meter antennas. IMO, the addition of a physical matching device to some of the other models of CB antennas...might also make some difference similar to what is noted for this model.
I posted this model on WWDX a while back, but it produced little interest.
This said, I tend to agree with DB when he says adding the matching feature to our models is beneficial to improved accuracy in modeling.
DB, can match his models using 4Nec2. If matching is essential for this idea why don't you ask him if he will do this project for you?