95% seems like a defensible value based on the MUMmer4 paper, which claims 98% identity across 96% of the genome. As I argued above, I think the 98% is probably an underestimate and should be 98.8%, which gives 95% as an overall value.
2 Likes