Resonant Shattering flares (RSFs) are bursts of gamma-rays expected to be triggered by tidal resonance of a neutron star (NS) during binary inspiral. They are strongly dependent on the magnetic field strength at the surface of the NS. By modelling these flares as being the result of multiple colliding relativistic shells launched during the resonance window, we find that the prompt non-thermal gamma-ray emission may have luminosity up to a few $\times10^{48}\text{ erg/s}$, and that a broad-band afterglow could be produced. We compute the expected rates of detectable RSFs using the BPASS population synthesis code, with different assumptions about the evolution of surface magnetic field strengths before merger. We find the rate of detectable RSFs to be $∼ 0.0001-5$ per year for BHNS mergers and $∼ 0.0005-25$ per year for NSNS mergers, with the lower bound corresponding to surface-field decay consistent with magneto-thermal evolution in purely crustal fields, while the upper bounds are for systems which have longer-lived surface magnetic fields supported by flux frozen into the superconducting core. If some of the observed SGRB precursor flares are indeed RSFs, this suggests the presence of a longer-lived surface field for some fraction of the neutron star population, and that we could expect RSFs to be the most common detectable EM counterpart to GW detections of BHNS mergers. The non-detection of a RSF prior to GRB170817A provides an upper bound on the magnetic fields of the progenitor NSs of $B_{\rm surf}∼ 10^{13.5} \text{ G}$.