The best is to connect something on the ODB port of the car and collect all the technical values.
You need to remember that the battery temperature can influence a lot.
The temperature does influence a lot, but it only influences within a range that's allowed by the current health status of the battery.
In my mind, this status is what should be somehow visible / easy to check. As opposed to trying to estimate it given the current temperature and the reported range when X% full or whatever.
My iPhone says its battery is X% healthy. I've seen my battery drain quicker when I was using the phone intensively by 0 °C than when it sat on my desk doing nothing at room temperature. But the health indicator didn't go up while in the office.
A full "coulomb count" of a battery can somewhat reliably estimate the amount that a Li-Ion battery has worn out. Its a pretty simple idea:
1. On one charge-cycle... count every electron that enters the battery.
2. As its discharging... count every electron that leaves the battery.
Typical phones and laptops will perform a coulomb count whenever you go through charge / discharge cycles (and a full coulomb count will only occur if you go from 0% charge to 100% charge... and a full count is the most accurate).
However, this is an external "blackbox" estimate to the life of the battery. You don't really know the battery's condition. There's no chemical test involved where you're sampling the contaminants inside of the Li-ion cells.
Is this "good enough" for when you're purchasing a used car / computer / iphone or when you're thinking of taking that long trip with your already well-worn Tesla?
---------
I'm fairly certain that any accurate chemical test would destroy the cells, at least with our current level of technology.
As such, its not a question of "good enough". Its simply a question of "what is possible". Since nothing better is possible yet, Coulomb Counting is our best methodology.