When I see chances to improve it, I will. It's likely that, once NL reaches a final state as far as features go, I'll spend a bit of time specifically on optimizing the code (because at that point - optimizations can be done without having to worry about what impact they might have on the ability to add new features in the future).
However, currently, the only options are "use a faster PC" or "turn off any other option that might slow things down" (eg. high quality minimap - which is already a huge performance hit, but even more so in high-res mode). In terms of what counts as a "faster PC" - the two things that matter here are single-core CPU performance, and RAM speed. Having more cores won't help (except to the extent of "the core isn't shared with as many other apps"), having more RAM won't help (NL is 32-bit and almost every PC these days has 4GB or more RAM anyway), having a better GPU won't help (NL doesn't use it), having an SSD won't help (it will improve load times slightly, but won't affect in-game lag).
In an ideal world, I strip out GR32 (or at least, use it only for physics rather than graphics) and replace it with a hardware-accelerated graphics library. In practice, that's not a task I'm willing to take on until / unless I decide the rendering code needs a rewrite anyway.