What does the powerplant do for Jump?
Here's my interpretation of why a powerplant of Pn=Jn is required. It's not incompatible with the rules or existing explanations.
Per LBB5, a Jump Drive contains Jump Capacitors that hold 18*M*Jn EP. Also per LBB5, these capacitors need to be provided with 2*M*Jn EP over 1 or 2 turns (20-40 minutes). That leaves 16 EP*M*Jn unaccounted for, and these have to come from the "overclocked power plant" part of the Jump Drive.
Here's where my interpretation comes in. The Jump Drive, if it gets the maximum energy out of its powerful-but-inefficient fuel burn, would actually generate all 36 EP*M*Jn for the capacitors. However, the Jump Drive fusion reaction doesn't generate a stable or predictable amount of energy -- it can be up to 11% lower than its theoretical maximum output. The powerplant is there to make up the shortfall. If it can't do it in one turn (20 minutes), it has to run longer to finish "topping off" the capacitors.
You can't build a Jump Drive to have the additional capacity to make up that margin because it can't be controlled precisely enough -- it will inevitably overcharge the capacitors, destroying the ship.
That's the first part of the interpretation. The second part is what happens in Jumpspace. Jump, despite Humaniti's thousands of years of experience with it, is still not well understood according to canon. Experiments have found that the Jump Drive should have external power applied to maintain the Jump Bubble. This power is applied for 158.2 hours, at which point it is safe to allow the Jump Bubble to collapse. For a normal Jump this may happen immediately or take up to 33.6 hours to happen. The time distribution of Jump Bubble collapse approximates a bell curve centered on 168 hours from Jump initiation.*
While there have been rare instances of ships staying in Jumpspace for the expected week even after suffering a power plant failure immediately after Jump, the usual consequence is loss of the ship. The likelihood of a mishap increases the longer the Jump Drive is unpowered before the 158.2 hour mark.
Powering the Jump Drive for longer than 158.2 hours does not appear to have any effect on the timing of Jump Bubble collapse. Neither does providing more power to the Jump Drive.**
No, I don't have any idea how much power the Jump Drive needs to keep the Jump Bubble stable.
*For reasons unknown to science, this is actually a pyramid rather than a bell curve. (Science doesn't know that it's a modified 2D roll.
)
**Theoreticians suspect that if the rules of the Universe were slightly different, it would be possible to bias the timing of Jump Bubble collapse closer to precisely 168 hours by careful adjustments to the Jump Drive. (These theoreticians may or may not have access to rules published after 1981.)