It is at least somewhat likely that the test cells are so small, or have such low specific capacity that they do not properly represent the effects of heat generation during charging. If these test cells are practically only voltage limited because they can dissipate heat rapidly, you can calculate the resistance of a real cell like the A123 M1 dropping to not less than ~1.5 milliohms with the new material and few other necessary modifications. This would allow charging in a minute which they do mention in their practical examples, but would rule out repeatable charging in less than 20 seconds.

It would be difficult to leverage very efficient cells such as these at their peak power in pulse applications unless great care is taken in the designs to prevent hot-spots appearing. In any case, improvement in the ion mobility reduces the duration maximum power can be tolerated even though it increases the duration you can hold at any particular value of power.

Still a big improvement, even useful in the case of being limited to a nominal 36 volts, simply because sag is generally significant.