67 ppm

The first metric is your GM does not know what he is talking about.
The other thing that might be happening is if you buy into the 1.5 shift nonsense, some people think you count both tails of the distribution which would be 6.8 DPMO.
What it should be:
If you buy into the 1.5 shift – 6 sigma is 3.4 DPMO
If you don’t – 6 sigma is 2 DPBO (that’s two defects per billion)

It is six to the left and six to the right. The movement of the entire distribution to the left or right will leave only 4.5 on the side that moved toward a limit. Look up a z value of -4.5 in normsdist in Excel and it will return the famed 3.4 DPMO.
The 1.5? It was an observation of a good process engineer back in 1984 or 85. Many, many have tried to bring proof since then, most notably Dr. Mikel Harry, but its all nonsense. Processes move and processes also can get wider over time, but with a great understanding of the process and rational controls to detect changes at the source will prevent anything even close to 1.5. I have, and I know Mike Carnell has, always been able to do better. The self appointed pundit for the proof of the shift on this forum has also admited to have never done a six sigma implementation. You decide who to believe.