Natural gas extraction from shale formations has experienced a rapid growth in recent years, but the low recovery observed in many field operations demonstrates that the development of this energy resource is far from being optimal. The ambiguity in procedures that account for gas adsorption in Gas-in-Place calculations represents an important element of uncertainty. Here, we present a methodology to compute gas production curves based on quantities that are directly accessed experimentally, so as to correctly account for the usable pore-space in shale. We observe that adsorption does not necessarily sustain a larger gas production compared to a non-adsorbing reservoir with the same porosity. By analysing the entire production curve, from initial to abandonment pressure, we unravel the role of the excess adsorption isotherm in driving this behaviour. To evaluate scenarios of improved recovery by means of gas injection, we develop a proxy reservoir model that exploits the concept of Pressure Swing Adsorption used in industrial gas separation operations. The model has three stages (Injection/Soak/Production) and is used to compare scenarios with cyclic injection of CO2 or N2. The results show that partial pressure and competitive adsorption enhance gas production in complementary ways, and reveal the important trade-off between CH4 recovery and CO2 storage. In this context, this proxy model represents a useful to tool to explore strategies that optimise these quantities without compromising the purity of the produced stream, as the latter may introduce a heavy economic burden on the operation.