OK, I'm going to ask the question and hope I don't get flammed for it.
Although this question was already answered, I'll like to give a definitive reply, hoping it would be the last one.
One thing that we learned over the years is that, whatever AFCAD you might do, it will NEVER be to everyone's taste. Some expectations are just conflicting with each other. There are user that don't care at all the default ATC, because they fly online so, for them, any side-effect on the ATC created by the desire to make the AFCAD as close as possible to real world, doesn't matter to them. Other users, instead, fly *only* with the default ATC, and they would prefer a more standard and close to default behaviour. And, there are many users who don't even FLY the sim, they just WATCH AI for hours and, for them, there are other priorities, like smooth traffic flow, correct arline matching...
If you combine this, with the fact that everyone use a different AI package, sometimes even *combination* of them, like multiple package installed, one for airliners, one for GA, it's almost impossible to verify everything.
Parking sizes, you mention...that's quite messy because, there are still many AI models out there, that weren't even created for FS9, they are still around from FS2002 that didn't had the airline codes in the parkings so, there was some kind of common conventions in the model authors to use non-realistic wingspan sizes in the model files, in order to trick a specific plane to park at a specific place, by playing with parking radius in the AFCAD against wingspan sizes in the MDL files. Of course, this method doesn't make much sense anymore in FS9 and FSX, thanks to parking codes but, since there are still plenty of models made like that around, it might become very complex to appeal to everyone.
What we found is that, there's no way a "perfect" AFCAD can be done during developement. If we had to do it with the dedication needed, we would just have to stop scenery developement, and start playing with AI package collections, testing them together, etc.
The only solution that seems to work, is to let the most interested users, like Harpsi, who is helping us since we did EHAM for Cloud9, playing with the AFCAD, and exchanging feedback with other users, to progressively improve it, until it becomes something *most* ( "everyone", is just not possible ) users would find it a reasonable compromise, and that will become the standard AFCAD. And of course, the forum would be always open for users wanting to share more customized solutions, like multiple wind settings, etc...if they like to do so.
I think our solution is the best compromise, because a good AFCAD will eventually come. Harpsi is credited in our manuals and gets some minor compensation, we thank him for helping us, and he has always been happy of this.
In this market situation, the line between surviving and losing money is *very* thin, and we don't have much to spare, we can barely keep FSDT into operation by releasing 4 sceneries per year. If those would become just 3, because we had to spend 1 month extra for each one to create the elusive "commercial grade AFCAD", we might just not be able to continue as a commercial entity. And no, a better AFCAD on the relase day, will never increase sales so much to offset the loss of 1 major product per year...what it's important is that people knows that a good AFCAD will eventually arrive, as it was for ALL our previous 4 products.