EP and Traitor are the people to really talk to about ZTMs, but I've picked up enough from them that I feel comfortable commenting.
Unless the sysop has edited the warp breakdown or course length, all universes of a given size will have roughly the same number of warps.
The last .1-.2% (assuming your ZTM strategy is sound) will basically always be bubble weirdness. Its difficult to describe using straight text, but I'll do my best. Look at this diagram
GATE - sector1
|......./
sector2
Now say you're trying to plot out from sector1. GATE will be the first sector, so you avoid it and try again to plot out. This time you get a no paths found error. So the ZTM moves on. But what it didn't realize is that sector1 is NOT a de, it's in fact linked to sector2. But a standard ZTM will show sector1 and sector2 as being dead ends when they're actually 2ways in a small bubble.
There is no realistic way to get a 100% ZTM without having 100% explored. It would take at least as long as the rest of the ZTM (and maybe longer) to map that remaining .2%. The calcs and tests involved to do the plotting are tortuous.
Thats just one example, theres others, but if you're really that interested, take it up with Traitor or EP, heh. The really weird problems they deal with and the really weird ways they deal with them give me a headache, heh. Have some aspirin and plenty of free time before you start asking these kinds of questions
Promethius, I'm curious. What method are you using for your ZTM? Rev's method?
I have heard of another scripter that has a private ZTM that is 100% accurate, but I've never seen it.