Talk about seeing what you want! I think culling data from a paper without reading the paper is very dangerous. Your analysis shows (assuming you've done things correctly) that the zircons from La Virgen volcano to be indistinguishable from concordia or your 'model'. So, nothing can be said about favoring your model over that which was applied in this study. Zircons (cores and rims) were all plotted together for La Reforma and do not fit your 'concordia' or your model! I'm afraid you've assumed your model correct and can't see the forest for the trees. The data you used do not support your model in any convincing fashion. You even state this same fact and then tout your model as superior. It's not. I'm curious why you did not plot the Aguajito data. They fall on a straight line, but it does not regress through zero. So, for the 3 examples from this paper, your model performs as follows
Case 1: Good line through origin fits your simple regression.
Case 2: Poor line does not fit your model does not regress through zero.
Case 3: Good line, does not regress through zero and therefore does not fit your model.
Your comparison to concordia using these data has a fundamental flaw in logic (do you know what that is?). So, of the three analyses only one fits your model (but also fits your concordia curve, thus provides us with no hint that your model is superior). The other two do not fit. I'd say you are 0/3 or if I am being nice 1/3. Not a good start for your model.
Cheers
Joe Meert