There’s quite a lot that is true - and quite a lot that is false or omitted in that blog post. For example, conveniently neglecting that copper salts can be highly poisonous (which is one reason why all copper pans meant to be used with acidic food and/or relatively high temperatures are tinned, silvered or lined with other materials). Super-high thermal conductivity is rather pointless when the majority of the thermal mass and surface/volume is water (conductivity 0.6 W/m·K vs. 400 W/m·K for copper and 16 for steel), and it’s not good for energy dispersion - a non-insulated copper boiler will conduct heat well “outwards” as much as “inwards” (in fact more, as there is an endless supply of cool air…). Electrical conductivity is irrelevant for the use case. Copper does work- (and age-) harden much more than steel, although it is more thermally stable, and it’s very soft, so mechanically it’s not very durable (ever wonder why bronze was invented?).
Don’t get me wrong - copper is a fine material for a boiler, but it’s not perfect, just like stainless steel. I’d rather see an objective discussion of the merits of both, and a conclusion that for this and that trade-off we decided to go with X rather than Y, instead than a paean to the virtues of copper followed by ‘this wonder material is so expensive that we decided to use something else instead’ - without then saying what that “something else” is. Given the choices available, and the time spent rubbishing stainless steel (which is also relatively expensive, BTW: 316 is 80% of the cost of copper), I can only assume that the boilers will be in brass, or plastic.
TL; DR: Bah, humbug. 😉
LMSC The L being such a big brand, why they aren’t sourcing their own supplies and assembling like some others do.
I suspect they aren’t quite as big as their share-of-voice may make them seem. Production volumes will be in the few hundreds per year, which is probably not enough for them to get much customisation… especially since there doesn’t seem to be a lot of leverage of components across models