The Miocene aged fossil you discard as evidence is confirmed by B.P.J. Molloy and M. Eagle as C. laevigatus - so not a proto-ancestor either. Low sequence divergence (still BTW 13 base pairs (rbcl + ITS) between them which is HUGE) is typical of the family Corynocarpaceae for the markers used as well. I contend you have interpreted the trees in Wagstaff & Dawson incorrectly, as well as what their paper, and indeed Molloy's states. I have discussed your ideas with Dr(s) Rob Smissen, Brian Molloy and Peter Heenan, and Mr Murray Dawson - all agree your interpretation of the genetic and morphological data presented in their work is incorrect. They await with interest your ecological assessment - for which their are also many alternative explanations.