Jump where "target" was included by previous Jump, but "end" wasn't fails
See attachment of https://musescore.org/de/node/287371#comment-909866
Simplified example to prove initial investigation to follow.
Initial train of investigative thought:
When taking a jump, the algorithm searches through the repeatlist up unto that point, trying to locate the last occurence of the jump target, to ensure we're replaying the "final" playthrough of that attempt.
So it finds the jump target (segno), but it's most recent final playback is the one caused by the previous jump; starting from there, it can't relocate the Fine, as the DSalCoda didn't include that part.
So we probably have to patch the target lookup to take into account finding "multiple final passes" of the same section