What makes limited acceleration more economic? Are there efficiency losses that are dependent on the time it takes to convert x joule from electrochemical storage to inertia?
It is more about the shift points in the transmission. Your engine is most efficient at around 90% throttle and "low" rpm. Your car is more fuel efficient are highway speeds then city street speeds (aerodynamic factors are insignificant at low speeds). The result is that engineers can trade acceleration for efficiency by changing the shift points, and you don't realize that the throttle is nearly wide open since the car isn't accelerating very hard at the low rpms.
Of course the engineers who design eco mode probably know more about how to do this than I do. I know they mess with shift points, it wouldn't surprise me if they did other things too.
Some EVs have a transmission - while it isn't strictly required (which is why others do not) for best acceleration you need one. If there is a transmission this is a shift point.
Of course EVs have completely different operating parameters. The rest of my post was respect to a internal combustion engine.
Some EVs have a transmission, yes. My Volt has a planetary gear set in order to share traction with the ICE. But that's not a pure EV.
But what they don't generally have, unless they are conversions from ICE cars, is changing gear ratios. Torque is mostly constant across all RPMs in an EV, so there's not really a need for one.
You also kill the batteries in shorter amount time if you do not limit the acceleration. The habbit of pressing the accelaration to the bottom will drain lots of power from the battery. It also encourages reckless/sporty driving. Best way to do this is on passenger cars is to have an eco/sport switch. On a passenger bus or truck you don't need a sport switch, since those vehicles are meant to be driven responsibily.
Yes. Resistive losses in the motor and inverter are proportional to the current squared (I^2). Accelerating hard requires more current so 2x acceleration incurs 4x resistive loss. It's not a huge effect in most applications, but it does make a difference.
Of course the engineers who design eco mode probably know more about how to do this than I do. I know they mess with shift points, it wouldn't surprise me if they did other things too.