I'm very much with MFS and ICT here, and I really don't see a substantial civil .v. military split in this. Unless you are dealing with a very old fashioned non-AFCS non-reversible aircraft, there are huge safety implications to the T&E of mission equipment, and generally avionics and airframe are inseperable in the way they are evaluated.
There's always a sliding scale of risk, but on the whole I think I'd class (say) first stalling and first use of autoland as similarly right up there in the high risk area, whilst determination of climb performance, and determination of navigational equipment performance are both reasonably low-risk, with crewing and planning adjusted accordingly. But that's not an airframe .v. avionics split so far as I can see.
For that matter, I'm next week (Wx permitting!), flying a TCd aeroplane to obtain new data for development and validation of a simulator, including specifics of handling and performance during take-off and landing. Next job after that is in another TCd aeroplane, looking at previously undocumented handling qualities issues in relation to the causes of an accident which happened last year.
In both cases, I'm remaining within the TCd envelope, but...
Anybody care to argue that this isn't test flying, or that I shouldn't worry about the safety aspects of crewing and briefing on either of those tasks?
G