The preference for upper body restraint goes back many years.
The principal concern relates to head injuries in a mishap with a useful measure being the Head Injury Criterion (or HIC).
See
https://en.wikipedia.org/wiki/Head_injury_criterion for some basic discussion.
The use of upper body restraint doesn't necessarily prevent head impact but it reduces the velocity and likely damage due to impact.
On a side note, the other concern relates to sidefacing seats where the use of standard seat belts and harnesses is a bit deficient in respect of upper body restraint. The typical flight attendent style of harness is far better for sidefacing restraint and does a much better job of protecting the occupant.
As to the OP's question -
Anyone hazard a guess as to how much the engineering would costs involved to retro fit your average 4 seater GA bugsmasha? Surely the same risks would apply to occupants in the cattle class seats of your A320, 737 etc. My guess is that someone in that office is behind on their KPIs and had to come up with something - anything.
(a) the engineering costs should be at the lower end of the scale as the design work is pretty straightforward. The harness/belt assembly will be a reasonably predictable cost. The maintenance cost, in all probability, will be the major component depending on how much work needs to be done on the aircraft to create suitable hardpoints for the hardware attachments. The usual problem concern is the change in load paths and whether the extant structure is up to the job. That is to say, there are no simple answers until an appropriate design engineer has a good review of the structure on offer.
Keep in mind it is not just a matter of bolting in a harness to a rigid structure and flying off merrily. What one is trying to do is provide a useful energy absorbing system with an intent to reduce peak and average mishap head accelerations to more sustainable levels.
(b) so far as the heavies are concerned, of course the risks are similar in concept. Where there may be a difference lies in the extent of lethality inherent in the structures - hence the use of such workarounds as delethalisation foam assemblies. The overarching concern is the HIC assessment - there are various ways which one can pursue this goal.
(c) as for someone in a back office considering their KPIs, I think not. The subject is fairly well understood and such a recommendation (presumably as an education effort) is to be applauded.