from https://batteryuniversity.com/learn/article/types_of_lithium_ion
"High energy and power densities, as well as good life span, make NCA a candidate for EV powertrains. High cost and marginal safety are negatives."
Cathode material theory is different from actual products on the actual market. Actual cells have much more than just the cathode material, including flammable electrolyte, for example.
Claim about cost is similarly wrong, At about $300/kWh, LFP is almost twice as expensive as NCA as it stands; again, looking at actual production cells, not some theoretical raw material cost. This is due to price fixing in Chinese LFP market, the small size of said market, and the economies of scale of modern high-energy density cell types such as NCA.
LFP cathode also has potential thermal runaway, just the onset temperature is higher and released energy is smaller. Energy and power density of LFP cells is more than enough to bring the cells to runaway (somewhere around 300 degC IIRC) whenever the cells fail internally or are abused externally. The result is slighly less dramatic than with NCA but can nevertheless easily lead in total loss of property. A few youtube videos showing a single LFP cell shorted in open air and "only" releasing smoke and not fire does not mean they don't cause fire when packaged as larger packs in real-world installations. A modern NCA cell in similar abuse conditions typically does not release
even the smoke, because it has several layers of protection (missing from the typical LFP cell).
LFP was dead-end; NCA has seen all the development by the big players, including safety.
Numerous fires have occured in DIY LFP conversions.
Really, the key is to only use good brand cells. If you have/want to buy crap (which I don't have objection against), I wouldn't lull into the false sense of security of LFP, just accept crap is crap and may catch fire regardless of the claimed cathode chemistry. I will correct each time someone posts this "LFP is safer" crap on forums. I have seen enough evidence of the exact opposite.
The root problem is mixing battery science with consumer product selection. There is a long way from not understanding theory, to understanding it, and
then understanding the market. If you try to make engineering / buying choices based on science popularization sites like Battery University, you are guaranteed to go wrong.