When using the Beltstrike fuel consumption table, I was originally inclined to figure a week on the j-drive consumes exactly as much additional fuel as running the m-drive at the equivalent level for a week would. But, see below.
(Bear in mind that asteroid prospecting in Beltstrike involves a lot of free drifting and running endless sensor sweeps, only maneuvering on those rare occasions when there is an interesting reading worth flying over and checking out.)
There are three main variables I can see here: 1) average stellar density within a 36 psc radius, which determines likelihood of popping out in -- or at least close to -- a solar system, 2) the average interstellar density of skimmable brown dwarves in the same area (the existence of which was largely still speculative back in the 1980s -- fortunately current models estimate this number to be fairly high), and 3) The Big One -- how much power plant fuel will be available afterwards to sustain life support during the time needed to detect a fuel source (brown dwarf, Oort cloud comet, TNO-class planetoid, or what have you) and then fly what may be dozens, or even hundreds, of AUs to rendezvous with it.
The idea that the power plant might run out of fuel mid-Misjump -- or even during a normal Jump -- without causing the ship to immediately precipitate out of Jumpspace and back into normal space strongly implies that whatever contribution power plants make to a Jump, it does not actually require a lot of fuel.
My theory is that during Jump, the power plant regulates the energy from the jump capacitors back and forth to the j-drive. For this, the power plant must be operational and capable of functioning at a level least as high as the number of the concurrent Jump, but when supporting a Jump it is powered not by fuel but rather charge flowing in and out of the capacitors, as determined by the Jumpspace topology of the calculated Jump course. It is, in effect a hybrid system of sorts. Should the power plant run out of fuel while in Jumpspace, life support can be switched to the batteries (with non-essential personnel being put into low berths, if available), but the power plant can still hum along, regulating the Jump charge until it all dissipates when the Jump ends.
This is about the only way I can make Misjump work as written without being a TPK by fiat. The upside is that starships barely consume fuel while in Jumpspace, which gives them a fighting chance to come out of even a six-week Misjump with at least half a load of power plant fuel left. Then, as the old saw goes, "Subsequent events are up to the referee."
I think they broke misjump survival with the '81 rules and didn't notice.
1977 rules didn't need a powerplant for Jump, and every Jump was the same regardless of distance: 10% of the ship's tonnage in fuel based on capability rather than actual distance. A misjump meant you returned to normal space with a bit more than half your powerplant fuel load left, regardless of duration. (The first half was used to get you to Jump Limit outbound.) This lines up with your interpretation of the concept, and it's quite reasonable.
As I see it, it was probably expected to be modified by referee fiat after the roll for misjump, though it wasn't literally written that way. "Ship destroyed" was the
mandatory TPK result, and it was up to the ref to decide whether a misjump would be a TPK or not after that. Keep in mind that in the small-universe setting of the first three books the referee might only have generated a couple of subsectors, and a misjump could easily put the players' ship well off the map.
Which meant rolling up a new subsector for them to land in, and then up to three more(!) for the route back home. Want them to survive? Put a world where they misjump to, or bump them a hex or two off the straight-line course to a world that's already on the new map. Want a TPK 'cause they made one too many dumb choices that session and it's just not worth the repetitive stress injury you'll get from all the world-generation rolls? It's an empty hex; time to break out LBB1, blank character sheets, and the dice.
The "Jump needs a powerplant too" rule broke that, because needing a powerplant and its four weeks of fuel for the Jump Drive (not just the Maneuver Drive as in the 1977 rules) meant that the 30-day clock on the fuel endurance started when the ship took off,
and kept running through the Jump too. Now a misjump didn't mean always exiting with half the powerplant fuel allocation left, it meant two chances out of six that the misjump would last longer than the powerplant fuel.
The Pn=Jn rule feels like it was brought in from LBB5. Starships without powerplants seem out of place once you start tracking weapon and computer (ECM transmitter) power usage, even if you're not using energy points in the LBB2 design rules. Keeping with the spirit of LBB2:'77, maybe it should have been "Pn=manuever Gs, but never less than Pn=1". Either way, it probably didn't occur to them to consider how it affected misjumps.
The later discoveries of interstellar bodies and the implications of extensive Oort clouds didn't change anything, but they should have.