80 Wh/kg beats the 70 Wh/kg of the cells discussed in the article, which was the benchmark for your proposal. Voltage doesn't really matter at all for this; you

80 Wh/kg beats the 70 Wh/kg of the cells discussed in the article, which was the benchmark for your proposal. Voltage doesn't really matter at all for this; you just end up with a few more cells in series (though it's admittedly awkward if you're targeting ~56 V max as the top of SELV where most single-chip BMS ICs top out at 18s). And even though most common LTOs are only rated for 10C, they're rated to do that below 0°C, in which domain they absolutely crush the ratings of LFP cells -- for LFP, you end up sizing for temperature unless you're looking at summer riding only. Admittedly you're not going to match 40 C LFP and NMC pouch cells with a cylindrical cell, and I'm not aware of anyone currently making LTO pouch cells; but I do think it seems unlikely that the sodium cells that are the subject of this article bring any really new capabilities to the domain of ebikes.