Is the problem that you can have 1/2 a turntable or 1/2 a water tower for example, so you end up paying to maintain a lot of items that you only need for the steam trains, that cut the savings of the conversion to D&E?
Which is why, if you're phasing out steam traction in favour of something modern, the efficient way to do it is by regions. Switch a region wholesale from steam to diesel or electric, as appropriate. Move the best steam locomotives to another region where they've not been replaced yet. Scrap the worst ones from both regions. Repeat as required until all steam locomotives are gone from the network permanently.
Yes, all, and gone. Anyone proposing to run a 1920s train on a modern railway should be given the same response as someone proposing to fly a 1920s aeroplane in and out of Heathrow. Just because 4472 is the same size as a modern train, doesn't mean it's any more up-to-date than an H.P.42.
I would say that selling off all the land piecemeal was a poor idea. Keeping route formations intact and actually looking ahead to the possibility of population growth would serve a lot of places today quite well. Passing them on to councils for bridleways and cycle routes would be nice - at the very least mothballing the lines and stations would be a start.
This gets into the question of what we mean by a 'sanity option'. British Railways was fundamentally expected to run the railway at a profit year-on-year, with the profit being returned to the owner - i.e., the government. Planning for population growth wasn't profitable. Gifting land to local authorities wasn't profitable. Selling it off was profitable - and just as importantly, got the Board off the hook for maintaining it. In some cases they'd probably be happy to sell it for a pound, just to get rid of maintenance costs, in preference to mothballing anything.
If you want to change that, you need to change the attitude around what a nationalised railway is supposed to do from Day One. Rail subsidies weren't accepted as necessary until 1968, and even then only where British Railways couldn't be expected to cross-subsidise a particular route from its profits.
So what would I do? First off, recognise that nationalisation is coming. Plan for it from 1943 or so - keep the Railways Executive Committe going. Publish a Modernisation Plan in 1948 or 1949. Nothing in OTL's 1955 Plan was news. Dieselise low traffic areas. Electrify the busiest lines - it'll take a few years to get going, so go for 25 kV AC off the bat. Bring in railcars where practical and appropriate. Rationalise services - get an Operational Research expert in to do what Beeching did but in the early 1950s, and with full data. Bring in liner trains, intercity services, and all that jazz. Shut all the useless lines as soon as practical, and get rid of them properly. Use the evidence from the OR report to target investment in the most productive areas, and demonstrate to the government that you're spending their money well. There's a lot to unpack here. I could (and perhaps should) do a full-fledged TL on it.