Beagle, you are of course correct, but so too is the essential point. If the data presented is true, then there has been a breakdown of the system integration and safety management processes, whereby the physical safety of each aircraft system may be safe in isolation, but functionally unsafe when mated and used.
However, EdSett is not wrong as for some years now (in my experience) Trials Installation aircraft have often acted as the sole Integration Rigs, for a variety of reasons. (Boscombe, especially, rightly complain about this). Primarily, as a means by which prime contractors can transfer risk back to MoD. That is, they claim “It works on the bench, so it’ll work on the aircraft”. They claim payment from a gullible non-engineer project manager and later, when the problem appears, stick their hand out saying “You paid up, our job was complete, we need more if you want it to actually work”.
This was formally condoned by 2 and 4 Star in DPA, and confirmed under FoI.