Jimrtex introduced this under the Mathismander thread. If anything, the most interesting revelation is that Kennedy seems interested in a dual standard that would require minimizing variations in both CVAP and total population between districts, subject to some maximum deviation threshold. A dual constraint gets the most fascinating with Section 2 VRA districts, because satisfying both Section 2 obligations and a dual population/CVAP standard simultaneously would basically require packing as many elderly white areas as possible into the Section 2 district while maintaining the bare minimum threshold for the target minority. I'd love to see you try making a map in a smallish state with a section 2 district under say a 10% maximum deviation in both total pop and CVAP and see what happens.
Section 2 with Hispanic CD's already uses CHVAP in general. What is interesting is if Kennedy's idea of having a maximum variance test becomes law, more states might start allowing such variances to reduce chops, like we are doing now with Muon2's scheme.