The price of lithium-ion batteries has declined by 97% in the last three decades

The realities of batteries is that most of the cells will be at 90% capacity while a few will have dropped to 50%. So recycling really can be just plucking out the good cells and breaking down the few old cells.

Those 90% cells will also have a pretty nice degradation curve as most of the loss happens within the first few years of operating.

Yes, performance of a battery declined over time, just as "battery health" declines on your phone today, for the same reasons. Mostly due to surface oxidation of the electrodes iirc.

Lithium-ion was never a good idea due to the fire hazard (in a crash they would just straight up explode, not even just cause a raging uncontrollable fire like LiPo, just boom) and Lithium-polymer could not support the necessary discharge rates (which I believe was also true of li-ion at the time too...). Early LiPo packs were only rated to like 0.5C or 1C peak discharge - this means, for a 2 amp-hour pack, 1C would be a discharge rate of 2 amps. Even then the packs did not really do well at their rated discharge rates - NiMH packs were far more tolerant of abuse in terms of discharge rates, in some cases you could push well beyond the rated spec and they would take it for a year or two's worth of sessions.

I was an "early adopter" on electric R/C airplanes back in the day (this would have been like 2000-2004) and the early stuff I used was mostly repurposed stuff from electric cars. The changeover to 4/5 or 2/3rds sub-C cells was a big deal at the time, the first "good" packs I owned were something called "HeCells" around 1800mAH or 2000 mAH, I also used KAN 1100mah cells. What I finally settled on at the time were these Sanyo "1950FAUP" cells which were extraordinary performers - we could push those cells to 70-80 amps peak discharge and they would take it. So in terms of the way LiPos are rated today[0], that is 40C discharge. This would only be doable for short periods, but it is necessary for launch and climbout (when the motor is running flat-out and the wing is basically stalled). Without the ability to sustain those currents it becomes very difficult to launch. Even still we did not "taxi" them like a normal airplane (or a normal model airplane), just wasn't enough power to get off the ground without a paved strip.

We typically expected about 10-15 minutes of flight time if you were flying normally, maybe 8-10 if you were really hotdogging it. So that means we averaged about 6C over the course of a flight (higher during climbout, less during cruising).

Another unsung miracle is how much motor technology has improved over the same period. My early motors were, again, repurposed brushed motors from Graupner's electric car lineup. We later moved onto some brushed Kyosho "magnetic mayhem" car motors that could push even higher levels of current. There was a Czech company called "Mega" who was an early mover on small neodymium brushless motors and that was a big deal at the time, they were far more powerful and far more efficient. Nowadays you don't even think about it, they're just a commodity, but at the time even a small brushless motor would run you about $130 or so!

[0] (tangent but it's kinda weird that NiCd and NiMH cells have traditionally been specced in amps but LiPo cells are specced in terms of "C discharge", obviously it's trivial to move between the two if you know the capacity of the pack but there's a definite difference in "style" of how the discharge rates are represented here...)

We needed long battery life - 3 years for a small 8051-based board with a data radio (kind of like wifi in 1989) that turned on every second or so. We created a power control ASIC that kept the 8051 off most of the time - it would wake up when data was available for processing. This was a proof of concept project for the military not a consumer item.

Most technology follows an S curve. It starts out slow with initial experimentation until most of the dead ends and pitfalls are eliminated, then there is a period of rapid growth as the tech matures and gains mass market acceptance, then it levels off with all of the easy gains mined out and you see only incremental improvements moving forward.

The middle part of the S curve is when you see the dumbest articles about the technology, with people extrapolating the growth linearly and talking about how it will rule the world in a few years.

Bingo, it’s the nicotine vapor