fsfox, if you attach your AD3 file to a post or e-mail it to me (don at stuff4fs dot com), I'll take a look.
Re the test file you sent me, I duplicated your results exactly. However, I received a host of coordinates under the error message heading "The ground poly object(s) bounded as follows cannot be triangulated". Those coordinates correspond to the missing ground poly objects.
Did you not receive those errors?
Those errors are due to the method of 100m. partitioning for FSX. But these were very complicate test objects - unlikely to be encountered at a simulated airport (i.e., lines crossing partitioning boundaries two or more times). Given the difficulty I've encountered developing an otherwise-acceptable partitioning routine, I'm reluctant to "open it up" again - at least until it is demonstrated that it presents impractical limitations with realistic GP objects.
From your screenshots, it does not appear that the missing lines are due to this issue. If you send me the AD3 file
for the airport in your screenshot above, I'll try to determine the cause.
EDIT: Having taken a second look, one of the two missing objects could not have crossed a partitioning boundary twice. There may be more at play here than I initially thought. Update later.
Don