The revision V chips have two additional resolutions that can be set: TwelveBitV and FourteenBitV. Those are power optimized versions of TwelveBit and FourteenBit respectively. Currently using them is not intuitive because the slope() isn't calculated correctly.
The wrong calculation comes from the NumberOfBits trait implementation:

impl NumberOfBits for Resolution {
    fn number_of_bits(&self) -> u32 {
        match *self {
            Resolution::EightBit => 8,
            Resolution::TenBit => 10,
            Resolution::TwelveBit => 12,
            Resolution::FourteenBit => 14,
            _ => 16,
        }
    }
}

The TwelveBitV and FourteenBitV variants are handled by the wildcard and thus are wrongly assigned 16 bits. I think the two variants just need to be matched when the revision_v feature is enabled.