Yeah - I see that and agree, but it can bite you the other way as well.
It can work both ways really.
In the summer time, you charge to full and over that time it takes to burn that range, you are mainly conducting town driving, with no A/C running and are driving very carefully.
Your efficiency is great and the improvement in predicted range is high, due to this.
After the next charge, the GOM reports a predicted range based on that previous full charge of the battery, with careful driving etc including.
The predicted range is artificial inflated based on the historical data.
Your next trip is made in poor weather conditions and at motorway speeds with a strong head wind.
The car is now rapidly readjusting its prediction, based on the historical stored data.
Then it can leave the driver wondering why his predicted range ( at the start of the trip ) is suddenly decreasing much faster that the car had previously reported ?.
Now this CAN catch people unaware !
Both ways of doing this are simply a prediction, hence the word GOM.
I tracked my predicted range on our previous ZS EV Gen 1 over two years and almost 20,000 miles in a similar manner.
So, I can relate in a similar manner, what is roughly going with our Gen 1 model.
Enough to say that battery degradation was a real thing on our Gen 1 model.
Other owners have reported similar / more accurate SOH figures from their batteries, by using a OBD dongle and software to interrogate their SOH of their packs.
My “gut” method and simple man maths gets me close enough to know what is going on with my HV pack.
Depending on your usage case, I have now learnt that you do seriously need to build in a fair amount of extra ( fat ) capacity when choosing your EV battery size, this will truly offer some protection for both winter and future battery degradation, because there will be some !.