I have been thinking about the series hybrid and it may not be ideal for conventional aircraft configuration. The weight penalty is rather high and it needs to be accounted with wing area. It seems that best way to achieve more wing area is to make the plane a wing itself. Flying wing design ends up with large wing area very easily and this can be used to account for the weight penalty.
Therefore I am proposing now this series hybrid idea to flying wing instead. It would also save the long drive shafts and the associated problems which are in the Northrop early designs there.
The engine that drives the generator could reside in CG inside the wing and the electrical drive which is light could be distributed in the trailing edge to several motors and propellers.
This way also it would be possible to get lower disc loading for the same power for high altitude flight by distributing the power to several propellers which would be distributed in the trailing edge. This would work as alternative for using large propellers as these many props would move as much air as the two large props which would make the landing gear unbearably tall. These smaller props could also be inside the wake getting drag reduction benefit from the Goldschmied wake propeller idea but in a bit different form. These props would be easier to manufacture because of the lower power per prop and also smaller diameter for aeroelasticity considerations and it would also enable optimizing the prop planform to reynolds number on the rotation speed meaning very drastic taper ratio (very pointy blades with thick roots, and high curvature).
Interesting case example for poor power to weight ratio flying wing is Northrop N1M. 120 hp takeoff power for 1750 kg plane. That is enormously low power figure. The plane was upgraded later to a bit higher power, but it flew with that power, indicating that it would be realistic to design a rather heavy plane as a flying wing without needing to ending up using enormously big engines.