a puffed-up lithium cell that's about to explode. soft, squishy, and very flammable!

Cell quality is important. You want to be using known good quality cells like those from Panasonic, Samsung, LG, etc..

How you manage temperature, charge and discharge is also really important, dendrite growth can cause cell failure in time. Charge temperature is extremely important. So you want to make sure you're using a smart programmable BMS where you can set up all the protections properly. Ideally one with as many temperature probes as you can find, 4 is good, 8 is better and some will have that many.

Otherwise making sure nothing can short out internally is important too, but it sounds like you're putting some thought into that. Most critical IMO from what I've seen on pre-made battery packs is making sure your series banks are well insulated from each other and have no chance of vibration causing the cells to wear through their heatshrink and touch each other.

Whoa, there is a LOT to unpack here so I will do my best. Please understand as I'm an engineer by training I will always advocate towards caution rather than YOLO so I'm going to spell out a lot of doom and gloom so read at your own peril. I'm not trying to poop on your hobby/business.

Inquiries; What kind of cells, 18650s? What brand cells do you buy and do you know their use history? Do you mix model numbers? I assume these are bare cells with no protection circuitry and you add the 5A fuse? What kind of fuse, a PPTC? And is that a typo, 64 in PARALLEL? 4S64P? That is bonkers, what the heck are you doing with them that requires that kind of amperage?

OK OK, so, do you use a BMS? How are you handling these packs other than the fuse? You NEED a BMS to at least balance the 4 series strings and monitor the packs temperature and shut it down if any part of the pack gets too hot. Lithium cells have a very narrow range where charging is allowed and a slightly wider range where discharge is allowed but at some point you need to turn it off. In terms of massively paralleling the batteries this is dangerous territory unless you are acutely aware of the cells, their limits, age and tight monitoring to notice when a handful of the cells are aging (getting higher ESR and forcing the others to handle the increased current which might exceed their safety limits). The end goal is to balance the current evenly between the cells in parallel and its not trivial, if you force all the current through a handful of cells you will ignite a lithium fire.

Regarding charging to 4V and discharging to 3V, most modern 18650 cells from good vendors charge up to 4.2V with a float charge so you're leaving charge behind but 3V seems way too low. I guess it depends on how long you want the pack to last. There is max and min voltages the cell vendor will recommend but then if you ask they can give you plots of the cycle life at different depths of discharge and typically in my world we don't go much lower than 3.2V to ensure the packs last well past 500 cycles. The deeper you discharge the more stress you put on the cathode.

Finally, loose wires are certainly a cause for fires and since you're hand making these its likely your biggest risk but in addition to that dendrite growth in the cells. If you abuse the battery either with temperature, over current/current imbalance you slowly introduce dendrites which keeps increasing the chance the cell will catastrophically fail with continued cycling. Its statistics and with the number of batteries you're playing with any little bit of abuse will grossly magnify the chance of a big failure to happen. Cells can and do spontaneously fail from internal damage but its a lot more likely to happen during charging.

In summary, you are playing with some serious power levels here, I would read up some more before you get much further. Especially before you step up to 26V! I hope this helps!

For what its worth, cheap labor markets and China especially are perfectly capable of building safe products. Most high end brands make their stuff in China. And Chinese engineers are very capable of making safe products. The issue is there isn't motivation other than a foreign country with foreign engineers probing them and pushing them to make a product the right way. Otherwise they can drop ship it cheaply with no brand, no warranty, no risk and the only exchange that matters is the sale and then they're free of all obligation. Companies with a brand need to stand behind it and are at legal risk so the stakes are completely different.

Ton of progress?

I would refer to that as "research ongoing but no progress yet".

The big problem is the higher the energy density the greater the release of energy in a catastrophic failure. For example if you were to increase energy density ten fold, then the release of energy due to catastrophic damage to the battery would be ten fold. Materials aren't even in the consideration, it's going to be an explosion.

We'd certainly welcome an EV battery that weighs greatly less, but safety is always going to be a forefront issue only by the physics of energy storage.

Still there needs to be a lot of improvement. I think the weight issue is the outstanding one. Higher energy density and greater longevity would certainly be welcome, but I think safety will always be skirting the edge. It's not been a fast track toward these goals, but it's still come a good way. Unfortunately I expect lithium-ion technology to hit a wall sooner than later.

Home gas & electric installations are a lot safer today than they once were. They're actually a great example of how a dangerous power source can be made a lot safer through better choices and engineering.

The first gas ovens and heaters used coal gas, which is carbon monoxide and hydrogen. The monoxide meant that coal gas was highly poisonous to breathe; and both monoxide and hydrogen are odorless — so you might not know the oven was left on unlit and killing your family.

Modern natural gas is mostly methane, with an unpleasant scent added so we can detect leaks. Methane is unbreathable, but still a lot safer to have in your house than monoxide!

(To be clear, the monoxide was not a contaminant; it's a fuel. Coal gas burns into carbon dioxide and water, just like methane.)

So here, we got a better gas technology: in fact, we got a whole different gas; a fundamentally safer-for-humans one — and on top of that, added an extra safety mechanism in the form of an odorizer, making the gas stinky instead of odorless.

Same goes for electricity. Modern home electrical outlets are grounded, greatly reducing the chance of a dangerous shock due to a wiring problem or a defective appliance. And modern circuit breakers are much more reliable than the fuses they replaced; and they don't break, so the homeowner doesn't have an incentive to defeat the safety device the way folks often did with fuses.

And modern kitchen & bathroom outlets have GFCIs, so an appliance shorting out through water (or your body) will break the circuit.

So here, we've gotten better electrical technology too: grounded outlets, circuit breakers, and GFCIs all make electricity much safer than when it was first installed in homes.

To put it morbidly: Coal-gas ovens and ungrounded electrical appliances were both once common means of suicide. Their modern replacements are safe enough that they don't work for that anymore. Same goes for cars, by the way. Modern gasoline cars don't put out enough monoxide to kill a person rapidly by inhaling the exhaust.

So yeah, I think it's perfectly reasonable to say we need a better battery technology, ideally one that doesn't do this sort of thing.

It seems to me that one simple (and therefore probably wrong) solution to e-waste would be to have retailers, distributors, and manufacturers accept nonworking or obsolete devices for responsible disposal or recycling.

Turn the one-way supply chain bidirectional. If you sell it, you've gotta take it back when it's broken. If you have an assembly line for making it, you can make a disassembly line for taking it apart. If that's expensive, well, apply some ingenuity to making it cheaper.

Want to deal with fewer broken devices? Make devices that don't break so much.