Sure, the point really is that, if drivers are more flexible in which trains they can drive, you can get away with fewer spare staff to cover absences etc. I'm sure that's true to some extent, but I find it hard to believe it would make much of a dent in railway costs. A quick check reveals staff costs form 25% of rail costs. At best I'd guess you might save 5-10% of staff through these kinds of efficiencies. That's not a huge difference. And there are other problems. For example, to take your example, I'm guessing right now that XC drivers in the West Country are trained only to drive Voyagers - because those are the only units XC uses there. If you want them to be more flexible, you'll need to train them to drive IETs too. Plus whatever trains GWR uses there for local routes. And you'll still be restricted to which routes drivers sign. Those training costs will add up. By the time you've taken all that into account, your savings are going to be a pretty small % of the railway's costs. And to gain those savings, you've potentially sacrificed any efficiencies that might come from competition, and from smaller companies being able to innovate more (although I'd concede it would be very hard to quantify those efficiencies, and they also may well be small).
Again I think you're being over-optimistic. I would imagine the prime determinant of how many customer service staff you need is: How many customers do you have, and how often do those customers raise complaints etc. That number is not going to change just because you have fewer TOCs. Again, you may reduce the numbers of managers you need, but how significant will that be?
I'm not sure that's true. My own view is based on a recognition that there are both benefits and dis-benefits of privatisation, combined with a sense that change works best when done gradually. We saw when the railways were privatised how damaging it is to suddenly inflict an enormous change to how an industry is structured, and I think there's a real risk that nationalisation would repeat that mistake. It's pretty clear that the current franchise model is no longer fit for purpose, and needs significant improvement, but I'm pretty sure that needs to come via gradual change, being guided by what is shown in practice to work and not to work. I'm also somewhat concerned that many of those politicians arguing for nationalisation are doing so for ideological reasons (they instinctively feel that public ownership is how society ought to be organised), rather than based on any pragmatic assessment of how best to run the railways (just as the Tories privatised the railways for ideological rather than pragmatic reasons 25 years ago).