Here are some observations from my own experiences, and those coupled with a good friend.
In short, AMS 2750 has proven to be a valid specification. In general, it seems to be fairly well written and concise. The problems encountered are NADCAP audits with AMS 2750.
Much of the problem comes from trying to integrate AMS 2750 with other industry specifications. If you were to just audit against the one specification, it would be all right. But trying to get everyone in the camp to agree on one thing is exceedingly difficult.
There becomes small differences between them; like 6 months vs. 1 year, type of process equipment, uniformity requirement, etc. Most of the auditors seem to be by-the-book, box checking auditors; with little/no process/risk-based auditing approach. Again, they check boxes and write NC's against the boxes. Instead of viewing the process as a whole and the entire myriad of specifications that heat treat facilities must deal with, it becomes a very frustrating ordeal of trying to find the most stringent requirement in print and applying it. Sometimes that is quite expensive/time consuming, and there is no process/quality/ statistical justification except "well it says it right there; you need to do it". That is not the intent of a good quality system.
In the old days, one could meet AMS 2750 and most "customers" (Bell, Boeing, McDonnell Douglas, etc.) would be fine with that, even if there were deviations from the specific company specifications.