Mr. Peterson can't seem to analyze worth crap.
Basically, his analysis boils down to... given a 24 kWh BEV (Leaf) that goes 73 miles that replaces an ICE version (Versa) , a 60 kWh BEV would have a higher battery manufacturing emissions and therefore, reduces the emissions advantage by 16%. Predictably, a 90 kWh battery would have higher manufacturing emissions and the advantage is even less.
Of course, that's all nonsense. First of all, 24 kWh BEVs have a limited use case. There are plenty of vehicles that would not be replaced by a 24 kWh Nissan Leaf. So, the comparison between a 60 kWh BEV or a 90 kWh BEV is not with a 24 kWh BEV, but with the ICE vehicle that is getting replaced, which isn't a Nissan Versa. A Model X doesn't replace a Versa, it replaces an Escalade, or a GL 550.
Furthermore, Tesla's cells are not made in China (as assumed in the research paper he based his figures from) and the vehicles are not made in Detroit. Additionally, Tesla's battery plant will have much better emissions than as modeled, so his input assumptions are pretty much garbage. As usual.