I don't think I've ever seen such a description of the variability.
It is implied strongly with the rules in TNE, as a crit on navigation results in a "bingo jump" - one that drops one into a zero thrust to orbit exit point. That's an event which requires foreknowledge of jump duration in order to know where to place the exit point. That said, for those who consider each edition a separate canon, it's ignorable. Marc isn't exactly one of those people.
For example, Earth's orbital speed googles as about 30 km/s... There are several variable durations listed in canon... And let's use AC-Earth as our baseline... someone on reddit puts it at 24.8 km/s. S... speed is about 54.8 km/s, since A/B Cen are not on the eccliptic nor parallel motion... but hyp gives about 40 km/s or 144,000 km/hr
Time | Source | T Min | TMax | Dev |
168 ± 10% hours | CT | 151:12:00 | 184:48 | ±2,429,200 km |
1d6: 1=6 days, 2-5=7 days, 6=8 days | MT Imp. Encyclopedia | 144 | 192 | ±3,456,000 km |
120 + (2d6 × 6) hours | MT: SSOM | 132 | 192 | ±4,320,000 km |
148 + 6d6 hours | MGT 1E CRB | 154 | 184 | ±2,160,000 km |
100 diameters is 1,274,600 km; hitting a bingo jump is unlikely without knowing the duration. Not impossoble, but definitely really hard. Especially given different ecliptics.
No way could you know in advance, there is no communication between jump bubble and normal space, except the influence of the 100D gravity well. Gravity precipitation may take a few hours and a change in the jump drive behavior would give you that much notice.
If the jump field slowly & steadily loses charge while in J-Space, then a few minutes in, the strength drop can most likely be determined to within a few minutes for exit time... but it there's a relativity error, experienced time aboard may be different
MTJ (which is canon, & GDW owned the rights, per the notice in the books) gives rules for group jump, so the variability is not inherently individual - tho' group jumps have normal variability for the group, and a very reduced variability to each other.