Has anyone got any good guides to the technique with sensitivities and reproducibility numbers?

The BIACore website is an excellent resource. I'd done some very preliminary analyses, but did not find it especially suitable for my purposes. However, I do know of excllent work done in the field of signal transduction prtn interactions, and pathogen signalling performed in my dept, which were done in tandem with NMR soln studies, and genetic analyses respectively.

I have lots of experience playing with protein protein interactions. The systems are extremely robust and reproducible.