I'm dubious anout the relative merit of the conclusion reached in the 'study'.
OK, not so much the conclusion, but the appearance of data to back up what is known.
Yes a lighter mass will accelerate and brake with less energy input, but the producers of the study add a second variable by 'downsizing' their simulated engines used with their simulated lighter cars. (slides 16 and 17)
They then use 'downsized' engines to maintain "acceptable" acceleration rates, not match simulation acceleration rates. (slide 28)
Because of these inconsistencies, I can't accept the conclusion that weight accounts for the amount of fuel savings (which is less than 1% as steady top gear speed, despite smaller engines and lesser performance).