There's an intellectual case for a privatised railway subject to economic regulation along the lines of water or electricity. However, whatever model you choose, you need to make a fundamental choice: do you (i) support loss-making operations (ie, most of the rural network, and the commuter network off-peak) - and if so, how?, or (ii) let the market rip, in which case you need to make sure other modes pay for all of their externalities, and then let the devil take the hindmost.
The reality is that we are always going to do (i) and have an argument about the correct level of subsidy and structure, as (ii) is politically undeliverable and probably environmentally unsustainable. Once you get to that place, the question is then all about what is the most efficient method for doing this? Personally, I don't see that the current model is fit-for-purpose because the premium payments are in no way guaranteed (see: every failed franchise handed back), so I'd prefer to have a single national passenger operator (either nationalised or not), and I see very little value in open access. Freight I'm much more agnostic about - whatever we can do to maximise freight is fine with me, but an intergrated BR Mark II model doesn't seem to offer much.