Its not really much different polygon-wise than the AIA B747 and I can't recall much complaining about those models killing your framerates.
This is true, but the model that the aia 747 essentially replaced, the FSpainter model, had framerates a LOT worse than the aia model, so that's probably why people weren't complaining.
I did some framerate testing this morning, and here are the results which I'm sure you won't care for:

My testing site is in the Arizona desert, clear skies, about 105 models arranged so that all LODs are in the field of view. The model radius of all tested models has been changed so that they are all identical (for testing purposes). There are no other aircraft being displayed besides the model being tested.
I've seen someone elses more extensive testing results which show an even greater disparity between the AIG and AIA 757's framerates, so I'm not showing "worst case scenario" here.
While poly counts obviously play a part in the framerate performance, it's not the only factor in play. The LOD distances (the distance at which one LOD model shifts to another) also have an effect on fps. The TFS a346 has a quite a bit higher poly count than the AIG757, but still has better fps.
The AIG 757 model is first rate from a visual standpoint, it has all those 3rd generation AI features people want, reversers, XML flaps etc, compressing gear, etc, but it has 1st generation AI framerates.
Of course, you can claim I have an axe to grind, that I'm not a dis-interested observer, so I would encourage everyone to do their own FPS testing and share your results.
David R



