@swamidass I don’t think there is much evidence to suggest that homogenization is a general trend. In fact, Davidson et al (2005) questioned this assumption:
The occurrence of significant isotope variation among mineral phases in Holocene volcanic rocks questions a fundamental tenet in isochron geochronology—that the initial isotope composition of the analyzed phases is identical.
@davidson I’ve heard that argument before but creationists just love to respond to literally every argument that was ever made by an “evolutionist”. More specifically, I found this objection:
One might question why we do not have more isochrons with negative slopes if so many isochrons were caused by mixing. This depends on the nature of the samples that mix. It is not necessarily true that one will get the same number of negative as positive slopes. If I have a rock X with lots of uranium and lead daughter isotope, and rock Y with less of both (relative to non-radiogenic lead), then one will get an isochron with a positive slope. If rock X has lots of uranium and little daughter product, and rock Y has little uranium and lots of lead daughter product (relative to non-radiogenic lead), then one will get a negative slope. This last case may be very rare because of the relative concentrations of uranium and lead in crustal material and subducted oceanic plates. I note that there are some isochrons with negative slopes.
@T_aquaticus Good point.