Thanks, I didn't realise research had already been done on this. That said, the research does seem rather biased from my cursory glance, with a few outlandish claims made and the benefits clearly overstated. There is no way that a £30/year card giving 50% off (allowing everyone to pay child rate off peak!!) will be more profitable than a £20/year card giving 30% off. The price tipping point for the majority of the population for the majority of journeys will simply be the cost of making a journey by driving, and most of the time the current railcard rate of 33% seems right. Giving more of a discount won't encourage more people to switch from driving, it'll just result in more money being taken from those who do.
As the report authors admit, assuming that the rate of change of demand with respect to fares is constant causes it to overestimate demand increases wildly.