Precise quantification of endogenous protein-protein interactions across live cells would be a major boon to biology. Such precise measurement is theoretically possible with fluorescence lifetime imaging microscopy (FLIM) but requires first properly addressing multiple biological, instrumental, statistical, and photophysical challenges. We present a detailed investigation of the last three FLIM-specific challenges. Using an efficient, highly accurate analysis code for time-domain FLIM data that accounts for all significant instrumental artifacts (in part, through use of a parametrized model for the instrument response function) and is rigorously based on both conventional statistics (full lifetime histogram fitting by χ(2) minimization) and novel statistics (single pixel fitting of lifetime populations using "maximum fidelity"), we address multiple photophysical challenges, including the proper side-by-side statistical comparison of fluorophore monoexponentiality, the precise assessment of fluorophore lifetimes and lifetime photostability, and the determination of acceptor dark state fractions. Finally, we demonstrate the feasibility of precise measurement of the interacting fraction of a protein across live cells with FLIM.