I see where we went wrong. I misinterpreted your use of %PV as measurement variation as a % of Part Variation. The PV of 320% means that the variation of the parts in this study vary more than three times the tolerance. If this reflects your process, then it is not capable. %PV is not typically used to assess adequacy of the gauge. That is normally done using %Tolerance for an inspection gauge or %SV / ndc for a process control gauge.