I can see some reasons why LiFePO4 is still not a good replacement for lead-acid batteries in all conditions.
Firstly, charging a LiFePO4 battery in freezing sub-zero temperatures damages the battery. Electric vehicles have a heater built into the battery, heating it first, only then charging it. However, the heater must be built directly into the battery pack, increasing its price. Theoretically an engine and a battery located in the engine bay would heat it, only very slowly, so that if the trips are long enough, the battery temperature might increase to the level where charging is possible. However, at least a temperature sensor would be needed right at the battery, and if the trips are short, the battery never goes above zero temperature, which would allow charging. A user driving lots of short trips in -30C temperature might be very disappointed to find that the LiFePO4 battery is losing charge and never being charged back. Lead-acid does charge in sub-zero temperatures, but slowly especially if it's very cold like -30C, so deep discharge in very cold conditions is possible with lead-acid but not as likely as with LiFePO4.
Also, if a LiFePO4 battery is discharged in sub-zero temperatures, the only way to make the car run again is a jump starter (which might damage the battery due to the high charging currents, unless the leads are connected and disconnected very rapidly, maybe only for few seconds, in which case the damage could be limited). A low-cost battery charger won't do as it wouldn't have heating function.
Furthermore, for the typical cheap Chinese LiFePO4 batteries in the ~100Ah range, typically supported discharge currents are about 1C, meaning a 50Ah battery would support only 50A discharge. This isn't remotely there to start a big diesel engine in -30C temperature, and even if the diesel would somehow be magically already running, the 50A safe discharge current wouldn't be enough to run the electric power steering, which requires so much current it can't be supplied by the alternator alone. It is true that with good design, LiFePO4 can be made to accept even higher discharge currents (but not easily as they require a protection circuit and offering a high-current-capable protection circuit is costly), but it would increase the costs of the battery. However, a very cheap 70Ah lead-acid battery can easily supply 700-800 amperes of current. It does suffer from Peukert effect, but remember that Peukert effect only applies to continuous high discharge loads. It doesn't apply to intermittent discharges like starting the engine or turning the steering wheel, requiring electric power steering to function. In cars, almost all high loads are intermittent, not continuous.
Although LiFePO4 is lighterweight, it is not physically smaller. Generally in cars where there is a lot of heavy stuff like engine under the hood, having ~15-20kg in battery weight isn't a dealbreaker, if the battery is physically small. Generally size is more restricting factor than weight, and LiFePO4 doesn't offer size benefits, as lead is a very dense metal.
LiFePO4 can suffer from continuous charge more than lead-acid. Lead-acid batteries can be used in UPS systems with a single voltage setpoint, where it takes 5 years of continuous float charge to damage batteries. In cars, that have a less-than-10% capacity factor, this would translate to more than 50 years (however, car chargers use a higher voltage, damaging the battery faster than in 50 years, and calendar life effects play a role too in addition to float charging life effects). LiFePO4 would require more sophisticated charging algorithms, with at least a current sensor in the battery terminal, and a voltage sensor connected directly to the battery which wouldn't include resistive drop. In cars where many alternators don't have a separate sense wire, they would instantly damage LiFePO4. So many car alternators are not good enough for LiFePO4, although with good engineering they could be modified to support LiFePO4.
While cheap Chinese LiFePO4 costs match AGM, they don't match flooded batteries that are still used in cheap cars, yes, even in cars with electric power steering and start/stop technology turning the engine off in stoplights. In the price conscious car market, offering AGM in a cheap car would dissuade buyers from buying that car, and LiFePO4 would have the same problem.
Lead-acid batteries in cars work just fine. Regular usage, driving the car occasionally, keeps them topped up, avoiding sulfation, and most discharge cycles are rare and/or shallow, so even a 10-year lifetime can be expected from a quality battery in a quality car. They are heavy, but the engine with all its accessories is heavier.
