A mid-infrared (MIR) imager and spectrometer is being investigated for possible consideration for construction in the early operation of the Thirty Meter Telescope (TMT). Combined with adaptive optics for the MIR, the instrument will afford 15 times higher sensitivity (0.1mJy as 5 sigma detection in 1hour integration in the N-band imaging) and 4 times better spatial resolution (0.08'') at 10μm compared to 8m-class telescopes. In addition, its large light-gathering power allows high-dispersion spectroscopy in the MIR that will be unrivaled by any other facility. We, a collaborating team of Japanese and US MIR astronomers, have carefully considered the science drivers for the TMT MIR instrument. Such an instrument would offer both broad and potentially transformative science. Furthering the science cases for the MIRES1, where high-dispersion spectroscopy was emphasized, we discuss additional capabilities for the instrument drawn from the enlarged science cases. The science cases include broader areas of astronomical fields: star and planet formation, solar system bodies, evolved stars, interstellar medium (ISM), extragalaxies, and cosmology. Based on these science drivers, essential instrument capabilities and key enhancement are discussed (see the companion paper Tokunaga et al. 20102): specifically imaging, lowand high-spectral resolution modes, integral field spectroscopy, and polarimetry.