It is true for EVs the weight contribution is reduced (and work done overall for a car vs a bike doesn't scale directly by weight), although for regen I think the max you recuperate is ~30%. For the air part, you do more work simply due to significantly more frontal area (although at the speeds a bike travels, the air resistance doesn't play as big a role, and the Cd of a biker probably is a lot worse than a car).
Assume efficiency of 87%. You lose 13% of the chemical energy when accelerating, and 13% of the kinetic energy when regenerating for a "regen efficiency" of 76%. Charging might be slightly less efficient than discharging, but the difference is very small.
So regen removes around three quarters of the acceleration and elevation related disadvantage of the heavier vehicle.
The drag coefficient of a bicycle is very, very poor. This goes for motorcycles too. I have no exact numbers, but I would not be surprised to learn that a very streamlined car had a drag approaching that of a bicycle, even with its much larger frontal area. The Roadster isn't all that streamlined, so I believe you're right in this particular case.
However, at bike speeds, weight does count even when maintaining speed, given a significant part of the work is rolling resistance from tires.
Wow, the rolling resistance is far higher than I thought. You're absolutely right. Interesting. I was going to ask wether rolling resistance is linear with mass, but I checked it up myself, and it is (or rather, linear with the normal force, which amounts to the same thing).
This is what I'm getting at: Lots of people seem to think that if vehicle A is heavier than vehicle B, then A will use more energy. That is incorrect, drag plays a huge role. The frontal areas of the two vehicles may be the same, and the larger vehicle might actually have a lower drag coefficient. The lighter one will obviously be at an advantage in stop-and-go traffic, but with regen, the difference is reduced a lot. Mass is a poor indicator of efficiency in an EV.
This may seem to contradict earlier postings by me where I talk about wanting to conserve energy by coasting. But the drag coefficient is fixed, you can't do much about it besides buying another car or drive slowly, which is boring. The mass related losses, however, can be reduced by the driver, by trying to conserve momentum. You might be able to get 10% more range than you would have gotten if you didn't think about it.