Well, here is a cool potential answer.
Wertheim and colleagues revisited the age debate with the coronaviruses in their viewfinder, back in April's J Virology.
They looked at genome sequences from the 4 genera, focusing on areas that did not shows signs of having resulted from the mixing of different viruses (recombining) and applied more recently developed models to calculate the time to most recent common ancestor (tMRCA).
Prior to this study, the tMRCA of all four CoV genera was 10,141 years ago (based on Patrick Woo's analysis of all four genera using just one region, published in 2012). In another instance, Pyrc and colleagues noted that HCoV-NL63 diverged from its MRCA around 900 years ago (with a rate of mutation around 3x10-4 changes per nucleotide position per year) based on analysis of the spike protein (S) gene.
This new study concluded that CoVs, like other viruses (which they list in the discussion), appear to be an ancient line of viruses. The average tMRCA was calculated to be 293 million years ago-give or take. Dinosaurs died out 65-million years ago I think.
The authors speculate that this period might also parallel the evolution of the bat and avian host species.
I wonder what their ancestors were infecting before that?