That CNBC article points out one of the big sea changes in the fast recent jumps from a used car average of 3 years (for
decades prior to ~2017) to 4 years (prior to 2019) to 6 years is a big change by rental car companies to hold rental fleets longer. The biggest sea change for rental companies over the last few years has been a switch to a higher balance of BEVs in their fleets. It certainly seems to me to be causative, the used car age average is rising for the first time in my lifetime just as BEV ownership is rising doesn't
seem to be a correlative coincidence and more indirect evidence that the BEV first owner lifetime is higher than historic ICE averages. (Possibly
much higher given that 4 to 6.1 year jump in just one calendar year.)
The early EVs have passed their 10 year marks. Some are closer to 15 years. There's a very long tail of BEVs already on the road. The only models that statistically have shown "battery degradation" enough to remark on have been the early model years of the Tesla Model S and Nissan Leaf before both companies invested in active thermal management of their batteries. So far "to remark on" was "noticeable compared to factory spec", but most of those batteries remained in first use (as car batteries). I've heard of Nissan Leaf battery replacements as a mini-industry, but not due to battery degradation, due to massive jumps in density upgrading the early Leafs to larger range than their original spec. There doesn't seem to be much of a market yet for Tesla battery replacements and most of the numbers thrown around are speculation and/or misinformation.