I agree, capacity can be a good indicator to determine the battery health status, battery heat maybe another one.
To check my battery heath status, I was thinking about comparing the battery's factory/ design capacity with the full charge capacity or usable capacity if the latter can be measured?
For example, for a Bosch Powertube 625Wh, the design capacity is 16.7 Ah from manufacturer data. Over time, I suspect my full charge or usable capacity (in Ah) would decrease. So, calculating the delta between these two values could possibly indicate my battery health status.
Another thought, may be just compare the original battery voltage when charged at 100% (e.g 37.2V) with the actual voltage I get now when the battery is charged at 100% (e.g. 36 V) and see how the latter Voltage changes over time.
So here, (36 ÷ 37.2) x 100 would give me a current battery health status of 97% or a degradation of 3%. Does this make sense?
I doubt the final 100% charge voltage is a reliable indicator or battery degradation because it's dependent upon the programming of the BMS and would vary from manufacturer to manufacturer. Also, I doubt if the peak voltage is linear with degradation.
I think if the manufacturer doesn't include the ability to read consumption (for example in Wh) and percent of charge.
This is what I like about the way Tesla reports battery state of charge (including range remaining). The BMS reports the amount of electricity used since last charge and will tell you to the nearest percentage point what the state of charge is (from 0% to 100%). The latter is an estimate, of course, based on the BMS tracking of voltage and energy consumed and is designed to have a "fudge factor" at the bottom (0%) so you don't accidentally run the battery flat because the state of charge estimate had drifted due to multiple partial charge cycles in the middle of the battery capacity.
Because the range remaining is based strictly on the EPA consumption rating, it's as reliable of an indicator of how much energy the battery can hold as is the percent charge. The range remaining is locked into the simple formula of energy remaining divided by Wh/mile of EPA consumption. This means if the car, when new, had 330 miles of range, and you charge it to 100% and it reports 327 miles of range, then you have less than 1% battery degradation.
Of course, the degradation of a Tesla battery is slower and much more predictable than that of an ebike battery because it has far more sophisticated temperature monitoring and regulation and the BMS is more sophisticated as well. But we could monitor the state of health of our ebike batteries much more closely if their manufacture used the same type of energy reporting as the Tesla uses for their cars. Instead, we get vague and highly variable estimates of range based upon whether the bike recently went up or down hill (amongst other variables).
I don't think there is an easy solution other than hacking the BMS system so you can peek under the hood or modifying the charge cord and installing a current measuring device on it. The latter is probably your best option. The most accurate measurement would require running the battery to zero and then measuring the amount of energy added energy by the time it had stopped charging. This figure would be very close to the total capacity but it's far from ideal in that it requires the complete discharge of the battery to start with.
I chose to not worry about battery health, beyond unplugging the charger before it finishes charging all the way (unless I will be going on a long ride that day). I have an eMTB and an eGravel and both batteries might outlast me!