Admittedly 50% is optimistic (hence 'hoping...'), as is the suggestion that such improvement would coincide [on a timeline] with a 150kw battery and superdupercharging....but still, there's plenty to be gained with future technology. Plus, incremental efficiency improvements are more than just additive.
For sure, mass drop from batteries (or future storage gizmos) will be huge. As technology improves we will see significant mass drop from alternative construction materials (i3-esque, composite wheels/seats, etc.) and designs (like crushable vs foldable crash structures). I think we may eventually see elimination of the mechanical braking system completely, which will be another big % drop. Downsizing/reinventing/eliminating the 12v system is another chunk (mostly from the battery). With all that mass loss, drive train components (motors, gearboxes, wheels, tires) can be downsized as a result, further reducing mass (AND permitting better packaging and thus better aero). A 150kwh Tesla under 2000 pounds is inevitable. That will be a HUGE efficiency bump from our 5000 pound MSes.
Aero really has plenty to go. There's obvious multi % benefits with things like mirrors and wheels, but new concepts and computing jumps will continue to add improvement as well--as much or more than the obvious big ticket items. Future manufacturing techniques will allow for finer shaping on the aero surfaces and fewer/tighter joints between components. Packaging concepts will evolve as designers move away from traditional ICE-based logic.
While high, electrical efficiencies can still be improved. We're basically driving around 5+ year old technology for many of the components, which is more or less only second gen EV tech (roadster = first gen). Improved thermal management falls in this category. There's % to be gained here, and any % gained means fewer batteries required...which means less recurring cost to the manufacturer...so its not like this stone will go unturned.
