Some good questions here.
A lot of the functionality of the control input devices are part of the overall design philosophy for the FBW sytems.
Once the decision is made is to go pure FBW, then the human inputs implemented are a composite of the participating pilots in the design and the "dreams" of the engineers that think they can have a system that improves safety, reduces pilot/crew workload, accomplish the mission, etc. Sorry if I appear to have any criticism of the "engineers", but this opinion comes from a pilot that flew the first operational FBW system in USAF. Maybe Concrode prototypes were already flying, dunno. I only know that the jet should fly about as can be compared to what we were used to and what the average light plane pilot would expect. Basic flight laws that Wilbur and Orville figured out over a hundred years ago.
Several reasons not to have mechanical connections between the controls in the family models.
Imagine a failure that only allows one control stick to have the input to HAL.
Imagine a mechanical failure that locks both sticks!!!!
Imagine a situation where the troop who has a good awareness of what the jet is doing and the other troop is clueless. Think AF447.
Imagine a mechanical failure between the two controls that does not reflect what the other troop is doing.
And the beat goes on.
All I know for sure is that we never had a problem with the Viper implementation since 1973.