In post #1 it lays the blame for MRA4 failure at the feet of the project manager...I'm a project manager for the company in question although nothing to do with that particular programme...It's not the job of the project manager to ultimately decide if a project should continue...My job is one akin to a conductor keeping the various parts of my orchestra in tune and on time...At various times in any of my projects I present a snapshot of where I am to people a deal higher than me up the food chain (review board/steering group) to decide if what we are doing still holds benefit, still meets the requirement, isn't being achieved by another project elsewhere (not likely in MR4 scenario!) and should it continue...as a PM it's not my decision to proceed and I abide by the boards decisions and react to their advice...it should be noted that the customer has representation at these reviews and must sign off on the results...So the MOD could at any time decide that the project isn't worth it anymore and close it with justification. Outside of these boards there are built in 'go/no-go' gates on the larger projects where in smaller reviews failures will show up...The production of the data for all of these reviews is produced by various people so it's not a case of one person being able to draw up the figures and hide damaging facts...
The detailed requirements for a complex project like the MRA4 must have been difficult to tie down...The devil is in this area and once you think you've got them they move around again...Once baselined they become configured and controlled items so it would be easy to see who was changing them and where the overspend was coming from if you had access to the change control documentation...