Originally proposed as a cosmological probe of the large-scale structure, line intensity mapping (LIM) also offers a unique window into the astrophysics of galaxy evolution. Adding to the astrophysical explorations of the LIM technique that have traditionally focused on small, nonlinear scales, we present a novel method to study the global star formation law using forthcoming data from large-scale baryonic acoustic oscillation (BAO) intensity mapping. Using the amplitude of the percent-level but scale-dependent bias induced by baryon fraction fluctuations on BAO scales, we show that combining auto- and cross-correlation power spectra of two (or more) LIM signals allows to probe the star formation law power index N. We examine the prospect for mapping Ha and [O III] lines across all scales, especially where imprints of the baryon fraction deviation exist, with space missions like SPHEREx. We show that although SPHEREx may only marginally probe N by accessing a modest number of large-scale modes in its 200 deg(2) deep survey, future infrared all-sky surveys reaching a comparable depth with an improved spectral resolution (R greater than or similar to 400) are likely to constrain N to a precision of 10%-30%, sufficient for distinguishing models with varying feedback assumptions, out to z similar to 4 using BAO intensity mapping. Leveraging this effect, large, cosmicvariance-limited LIM surveys in the far future can scrutinize the physical connection between galaxy evolution and the large-scale cosmological environment, while performing stringent tests of the standard cosmological model.