I think you’re right, which is why there is different between the relative Ns and the relative rates per base per generation in that table, but I’m not entirely clear on why. Something to do with the fact that 50% of transverstion types are strong to weak or weak to strong, while 100% of transitions types are maybe?
EDIT: never mind, ignore me. There’s no difference between the relative Ns and rates per base pair per generation after all.
The group is not public, and my interaction with him was actually limited to his comment quoted above (in response to my referencing the biologos article to someone else) and my brief reply. He has not responded.
But thank you for your responses here!