A salient point. Molecular divergence is more crudely estimated, but closer to the point of actual lineage divergence, than morphological divergence.
So if you accept a placentral reproduction system to be monophyletic (and I think there's good reason to do this), the data points more towards it being a Maastrichtian or later thing than being deep in the Mesozoic - although the *lineages* might have split that early.
Indeed, it might even be so that evolution of the placental reporduction system was *both* faciliated by the dinos' extinction, and one major factor as to why they couldn't make a comeback - a trait that made for superior better "comeback potential".
E.g. the case for survivorship of maybe a dozen Charadriiformes lineages is reasonable.