The main improvements seem to be in management electronics...
...widespread teething problems...
Mobile phone circuits and algorithms have been broken down (since forever) such that any function not in use can be switched off... that does not help massively with an e-bike as there is only one function.
Li is inherently dangerous, but by monitoring a battery's condition it is possible to charge it safely and faster than before... that is a huge plus for users.
Yes, tiny tweaks have been made to improve performance, but fundamentally it is still the same chemical reaction (and if you remember anything from school chemistry, you will know it is a miracle batteries work at all!)
My point is, over 20 years, the performance of electronics has improved 1,000 times, whereas battery performance has quadrupled, perhaps? So while technology is riding a wave of electronic advancement, batteries are a drag!
There are lots of very clever people trying to prove me wrong - in the car industry (for eg.) a tiny margin in performance is a BIG deal! But I’m still sceptical: for 70 years the electronics industry has known precisely how to achieve the next leap in performance, with batteries each tiny step makes it almost impossible to optimise any further in that direction.
You are certainly spot on in terms of improvement in battery capacity - there has been next to none in the 10 years I've been ebiking.
The various lithium technologies were a leap forward over lead acid.
My first lithium battery was a Bosch one in 2010 with a capacity of 400wh.
The current battery is 500wh, but the case is slightly bigger.
Not quite proportionately bigger, so there has been an improvement in energy density, but only a small one.
I started the thread in the audax section, so my interest is in going longer, not faster.
Range for distance use remains the weak point of ebike batteries.
To get anywhere near 100 miles it is still necessary to carry two or even three lumpy batteries.
Given the glacial progress of the last 10 years, it's unlikely my cycling career will last long enough to see any marked improvement, and I may not even live long enough to see it.
My girl has done 100 mile rides and 100km AAA events on her Ribble with just the internal battery with about 15% power left at the end.
The light assist ebikes such as the Ribble and Orbea Gain are a slightly different kettle of fish, although there's no magic about the battery and motor.
They give a lot less assistance by using a smaller motor set up to draw less current, hence less battery use and longer range.
My Bosch ebike would go a lot further on lowest Eco setting, but the problem is the assistance on Eco doesn't do a lot more than compensate for the extra weight of the bike, particularly when ridden by a lumpy rider.
Range does vary a lot, but the majority of ebikers have what we might call 'full assistance' ebikes which in real world use give a range of 50 miles or fewer per battery.
As regards batteries used for electric car propulsion, the big car firms making genuine efforts to increase range must have a positive overall impact on the technology.
Although I'm told the cells being developed for car use demand properties that are not ideal for use in an ebike battery pack.
Every now and again, a research body publishes details of a new battery which looks very promising for ebike use in their laboratory tests.
One of these may reach production, but most are never seen again.
This has happened so often that it's now a standing joke among those of us who follow the technology from our computer screens.
Another factor that limits advances in ebike technology is the market is far too small for anyone to want to throw large amounts of resources at it.