Ok, no one produces new assets, so we don't need to improve the engine. And as long as we keep the old engine, certainly no one produces new assets. Sounds like Chicken & Egg to me. So we need an engine that can use the old assets and run on the same old hardware, but can also use new/better assets on more potent hardware. Unfortunately that engine does not exist.
I think it's more of a case where the work and extended release cycle incurred by porting trem to a new engine isn't very appealing if the end result is not a substantial improvement over the current version. I really don't think that "If you build it they will come" applies here very much - aside from mapping, the core assets have remained almost unchanged throughout the 1.1 release.
@thorn: Massively parallel trem would be nice. Is IOQ3/XreaL multithreadable?
