> More stop and go with normal braking: -50% efficiency
Wrong value. Apart from the fact that the number is invented, because each stop turns all the car's kinetic energy into heat, the number should be -100%. Either that or perform a full analysis, with a listing of all the factors.
After you recover 20% of the energy with regenerative braking, you end up with -80% instead of -100%.
I ride a low-power motorcycle. On the freeway, I get about 50 mpg. On back roads, with braking and generally higher rpm, I get 60+ mpg. For my low-power bike, the drag at freeway speeds takes more effort to overcome than does the varying speed and braking of back road riding.
And that is without regenerative braking. If I had regen braking, I'd get even better economy on back roads.
But regenerative braking cannot produce net positive energy compared to not braking at all -- that violates basic physical principles.
Regenerative braking is better than no regenerative braking, but regenerative braking is not better than not braking at all. That has been the point all along.
> = Net city driving with regen braking: +10% efficiency
No, not compared to driving at a steady pace. If the scenario you describe were so, it would represent a gain in energy greater than 100%, and that is not possible.
Stop-and-go driving is always less efficient than driving at a steady pace at the same average speed. Regenerative braking can only minimize the loss caused by stop-and-go driving, it can't reverse it -- it can't produce a net gain as you are implying.