An ndc of 5 gives you 5 statistically significant categories of measurement (or true resolution when considering gage error)- just a little better than a gas gage in a car. It is OK for simple measurement, but inadequate for SPC - in spite of what the chart says on page 47 of the MSA 4th ed. For SPC, the ndc>10 (5 categories on either side of the mean), or you will get "chunky" data and marginally useful results, if that.
The parts you should present to the Gage R&R should represent the variation you expect that gage to see over time for that characteristic with that process. Otherwise, all of the data will be either overstated or understated.