Process Capability and Discrimination Ratios

During a recent phone call with Douglas Relyea, author of the book The Practical Application of the Process Capability Study: Evolving From Product Control to Process Control, he mentioned some common questions that arise when consulting with leaders or organizations. I'll let Doug take it from here:

"A company CEO asked me this question just this week, and I receive many similar questions in regard to gauges:

I understand the measurement process analysis (MPA) performed on product XYZ bond strength indicates we have a discrimination ratio (DR) of 2 which, if I understand correctly, means we can use this gauge to separate XYZ product into only two categories – good and bad. Should we buy a new gauge?

My answer is:

No. A low DR is not always a negative. A low DR can be the result of manufactured product that has very little variation as compared to the variation of the measurement process.

In this case, the product variation is 9 gms/in2 and the measurement process variation is 47 gms/in2. The customer specification is 400 gms/in2, minus 50 gms/in2 with no upper limit stipulated. The production records indicate the bond for this particular product is generally produced with a mean of 500 gms/in2. This measurement process is suitable to tell good product from bad and, in this case, that is all that is required. "

Do you agree with Doug's assessment?