How unpleasant it is to read above, I agree. There is little doubt that Super Charging is extremely challenging for the 18650 NCA cells, which are High Energy cells with thick and less porous layers as opposed to High Power cells with super open and thin layer of materials.
This thorough German paper from Peter Keil describes ways to ensure BEV Battery life of at least 1000 Equivalent Full Cycles. It even has a section on degradation (dominated by lithium plating) when solely doing Super Charging.
https://mediatum.ub.tum.de/doc/1355829/file.pdf
Super Charging is NOT one of the means to reach 1000 EFC:-( In fact SUC simulation tests were stopped after just 120 EFC when degradation reached almost 25% (Pages 98-101)
Because of the severe degradation rate, the test were repeated with a reduced boost protocol and as well with a maximum cell voltage off 4,1V. The paper reports, even with reduced boot protocol - a capacity fade of 20% already after about 300 cycles from 3.2 V to 4.2 V or 700 cycles from 3.2 V to 4.1 V.
300 EFC is less than 350 km x 300 EFC = 105.000km == 62.000 miles.
In this light, Tesla 'COULD' decide to take the approach, that Tesla Model S users, who dominantly DC/SuperCharge and regularly goes to 100%, need to be tought better practices. A warning early in the cars lifetime, about the likely effect of the usage would have been in place - after all Tesla modified the charge curve for S 90D dependent on charging history.
Testing the maximum Voltage achievable is largely an anti recovery test:-(
(Other papers (from Peter keil and others) documents various Capacity recovery mechanism's in Li-Ion cells. One is Lithium deposited in the Anode Overhang area, which CAN be recovered with long periods at low SoC, another is slow cycling to reduce lithium plating. A dV/dAh charging analysis can immediately show in-homogenous lithium distribution (dv/dAh peaks smeared out) which ask for rest and severely left moved anode peak, which argues plating. The Tesla BMS can perfectly well do dV/dAh analysis, but to my knowledge, Tesla can only measure voltage on each full string of 70+ parallel cells, which wipes out the peaks we want to find, so Tesla potentially has to decide strategy on other measures (such as DV versus AC charging ratio))