[Abridged] Observations of molecular gas at all redshifts are critical for measuring the cosmic evolution in molecular gas density and understanding the star-formation history of the Universe. The 12CO molecule (J=1-0 transition = 115.27 GHz) is the best proxy for extragalactic H2, which is the gas reservoir from which star formation occurs, and has been detected out to z~6. Typically, redshifted high-J lines are observed at mm-wavelengths, the most commonly targeted systems exhibiting high SFRs (e.g. submm galaxies), and far-IR-bright QSOs. While the most luminous objects are the most readily observed, detections of more typical galaxies with modest SFRs are essential for completing the picture. ALMA will be revolutionary in terms of increasing the detection rate and pushing the sensitivity limit down to include such galaxies, however the limited FoV when observing at such high frequencies makes it difficult to use ALMA for studies of the large-scale structure traced out by molecular gas in galaxies. This article introduces a strategy for a systematic search for molecular gas during the EoR (z~7 and above), capitalizing on the fact that the J=1-0 transition of 12CO enters the upper bands of cm-wave instruments at high-z. The FoV advantage gained by observing at such frequencies, coupled with modern broadband correlators allows significant cosmological volumes to be probed on reasonable timescales. In this article we present an overview of our future observing programme which has been awarded 6,500 hours as one of the Large Survey Projects for MeerKAT, the forthcoming South African SKA pathfinder instrument. Its large FoV and correlator bandwidth, and high-sensitivity provide unprecedented survey speed for such work. An existing astrophysical simulation is coupled with instrumental considerations to demonstrate the feasibility of such observations and predict detection rates.