Dave,
A major limitation is that existing variable power transmission devices, which would control this allocation, can consume a meaningful percentage of the engine's power.
Depending upon what your definition of "meaningful percentage" is, I might agree or disagree. Here's a
traction CVT that I've been working on. I did a design study for a split path MRGB variator sized for roughly 400 ft-lb at the variator input and a drive ratio of 2.0. The variator components would have added about 105 lbs to the MRGB. The variator input-to-output efficiency would have ranged between 93% and 95% during variable speed operation. With a traction drive, these losses are primarily due to EHL contact spin loss.
Being a split path configuration, only half the power passes through the variator. So worst case, the variable speed device itself would produce less than a 4% loss. Not as good as a gear system, but not too bad either.
The biggest issue with using a traction device like a CVT (ie. a power transfer device lacking positive engagement) in an aircraft drivetrain is how do you demonstrate that it has adequate fault tolerance/reliability?
Regards,
riff_raff