I believe there are some metamaterial-based approaches for the actual microbolometers that should help drive costs down. Ostensibly metamaterials could also be used for lenses but I'm not aware of anything there, we're probably stuck with germanium for the time being.