Wow, so I went and downloaded their "methodology" document. Some highlights: In prior years, they made an attempt to estimate the true fuel consumption and emi

Wow, so I went and downloaded their "methodology" document. Some highlights: In prior years, they made an attempt to estimate the true fuel consumption and emissions rate of gasoline cars (which is invariably higher than the EPA figures), but in this year's edition they reverted back to the EPA numbers because "To what extent [the prior year practice] is appropriate for new vehicles today is unclear". This makes gasoline cars and hybrids look better than they really are. They assume that PHEVs drive 50% of their miles on mains electricity. This is, needless to say, a completely bogus assumption. They randomly add 11 cents per kWh to the cost of nuclear power to account for possible future nuclear accidents. Okay, whatever. Their "green score" function is heavily skewed to mask the damage inflicted by gasoline cars. The difference between a score of 71 (Toyota Prius) and 57 (Tesla Model Y) is about 0.5 cents of environmental cost per mile. The difference between a score of 30 (GMC Sierra) and 20 (Mercedes AMG) is 1 cent of environmental cost per mile. In reality there is a larger difference between the (bad) GMC Sierra and the (even worse) Mercedes AMG than there is between the Prius and the Tesla, but their "green score" makes it seem like the difference between a Prius and a Tesla is larger.