While light tight oil reservoirs such as found in Bakken and Cardium style plays can be stimulated by fracturing, significant oil production from true shale reservoirs dominated by clay size fraction sediments may not be feasible, even when reservoirs are fractured, if current oil storage models in source rocks are valid. We look at the source rock knowledgebase and the gaps in our understanding relevant to assessing oil production from shale sections. Despite over four decades of study of petroleum source rocks by geochemists, our mechanistic understanding of them remains at best qualitative and inadequate for quantitative reservoir characterisation purposes. Past studies emphasised molecular chemistry details and gross mass balance studies at large scale, but only few detailed mechanistic studies of actual physical process and transport function during generation and primary migration are available. Empiricism and generality dominate current concepts and most R&D peaked in the 1980s when Oil Company R&D spending was at its highest, but declined after that. Today there are many areas of uncertainty, but reassuringly, research is starting again and crucially, abundant core samples, which were largely absent before, are now also available. There has also been a revolution in the understanding of shale sedimentology which forces us to reexamine many of our earlier precepts about source rock formation and suggests that in the absence of naturally fractured silica or carbonate intervals within source rocks, such as found in the Monterey Fm., very small scale collateral silt and sand horizons may be locally important small scale reservoirs in typical shales. The bulk of the oil storage in source rocks however is in kerogen, where dominantly diffusive rather than Darcy flow processes operate and where fracturing may have little impact on productivity. We look at the technical barriers, both theoretical and analytical, to understanding true shale dominated reservoir oil production potential.