Tesla’s Nov. 16 introduction of both the new Tesla Semi and the Tesla Roadster managed to impress even observers used to the electric carmaker’s lofty ambitions. When they come to market, the Roadster will be the fastest-accelerating production car in the world, and the Semi will deliver more range, for a lower price, than almost anybody expected.
Or so Tesla says.
According to a new analysis by Bloomberg, those and other promises would be difficult or impossible to fulfill with batteries on the market today. At current battery prices, Bloomberg estimates the battery for a freight truck with 500 miles of range would cost more than $100,000, but Tesla estimates that the entire truck will only cost $180,000.
Similarly, the power and range Tesla is promising for the Roadster will take a battery more than twice as big as anything in a current Model X or Model S. One analyst told Bloomberg that “I don’t think it’s physically possible to do that right now.”
Charging is another major question mark. Tesla says the Semi will be able to suck up 400 miles worth of juice in 30 minutes, but that would mean charging more than 10 times faster than Tesla’s current best chargers.
Get Data Sheet, Fortune’s technology newsletter.
But here’s the trick: the Semi isn’t set to hit the road until late 2019, and the new Roadster isn’t due until 2020. Battery prices and sizes have dropped dramatically in the last five years, and many projections have them dropping by another half by then. The same goes for power density, meaning that by the time the Roadster and Semi go into full-scale production, batteries may well be good enough to do what Elon Musk is promising. A Carnegie Mellon battery expert speaking to Jalopnik agreed that the Roadster’s performance benchmarks were at least theoretically plausible, though they might also require advances in tire technology.
Announcing products years before perfecting the technology needed to make them work is certainly an unorthodox approach. It’s unlikely to appease critics of Tesla’s rapid spending, who tend to see displays like the Roadster reveal as media stunts aimed at raising badly-needed cash.
Moreover, trendlines don’t create products, and Tesla will have to actually do much of the work to improve batteries while making them cheaper. It has already done a lot of that with the increased capacity brought online by its Nevada Gigafactory — but there’s a glitch there, too. It was recently revealed that battery production issues were a big part of what has slowed down production of the new Model 3 sedan. That has torpedoed Musk’s typically optimistic timetable for Model 3 delivery, while leading some to question whether Tesla is really the repository of battery genius it has been taken for.
In other words, it all comes down to execution. So far, Musk and company have managed to pull off most of the big challenges they’ve set themselves. But they’ve rarely managed to get across the finish line as quickly as they’d hoped.
