Kitbag, that is excellent stuff and all true, but with MoD falls at the first hurdle...
Strong but small project offices must be provided both by the military and industry.
If by some chance an MoD project office is strong, senior management will almost certainly not tolerate such an attitude and will definitely not support you.
From my own perspective, matters changed in 1996, following the appointment of Sir Robert Walmsley as Chief of Defence Procurement. His first action of note was to announce he was cutting 500 engineer posts at AbbeyWood, e-mailing everyone on site that he saw no need for engineers working on engineering projects or programmes. (Thanks Bob, we’d all just been posted from London. You could have screwed us before upping sticks).
A major prime contractor on an aircraft programme was a thoroughly unscrupulous outfit who enjoyed the personal patronage of CDP. Their boss kept a little black book of the strengths and weaknesses of the MoD team. He soon approached a non-engineer saying “Could you just confirm we have a test and declare status?” The MoD guy hadn’t a clue what this meant; the company knew he’d never admit this and seek advice. He said “Of course”. A couple of years later and “aircraft installation, systems integration and demonstration of installed performance” consisted of the prime pitching up at Westland and dumping a truck load of LRUs on the hangar floor. “We tested them on the bench and we declare we can’t tell if some work properly as they haven’t been integrated or flown”. They were paid in full, approved by CDP and the 2 Star (Nimrod MRA4, Chinook HC Mk3 etc, which I mention just to illustrate the quality of decision making here), and a second payment approved for Westland to meet the Prime’s original obligation.
By the way, this attitude of “it works on the bench, so it’ll work on the aircraft” was a significant factor in the loss of Tornado ZG710, which the same 2 Star (DGAS2) was personally warned of 4 years before the accident.