It's probably worth restating my philosophy here since it is buried under a dozen pages of postings since.
We are talking model railways here, so my purpose is to be able to run locos and carriages I like in colours I like. (Freight stock is a bit freer given a lot came from other companies through interworking and pooling arrangements). It's a bit of an antidote to the rivet counting philosophy that says A couldn't have run with B and C is the wrong colour for the time A was around, without going over to a completely anything goes approach.
Now on a good model railway geography does play a role in terms of the scenery, the architecture of buildings and in what goods are carried in trains, but since few of us can contemplate the sort of American basement layout where a whole network can be modelled, we are unlikely to model multiple stations. So I don't think it's necessary to plot the route of a line that was never built. There were very few viable routes left unfilled in Britain anyway, certainly not after the last gaps were filled in with light railways.
So my approach has been to take lines that were built and give them a different history. In the posting above I see @Regularity has a picture of Lydnam Heath, that junction where Bishop's Castle trains had to reverse. An alternative history might have been that the line to the Welsh coast was completed. Or, heading north a little, another alternative history is that what became Colonel Stephens' Shropshire and Montgomeryshire had been successfully completed in the mid 1800s and the vision of a mainline link between the Potteries and the Welsh coast via Shrewsbury had become a reality.
A large number of lines in the mid-nineteenth century were promoted by independent interests and only when completed taken over by a major company. An obvious alternative history is that those companies remained independent. This the approach I have taken. Instead of the Redhill, Guildford and Reading line becoming part of the South Eastern Railway I have it staying independent. Similarly the independently promoted lines between Leatherhead and Horsham and Epsom and Mitcham Junction didn't fall into the lap of the LBSCR.
As far as a layout goes, I am only thinking of a small model of one of the junctions near Dorking as a showcase for the stock I am building. Another possibility, but discarded as being too large and ambitious, is the terminus at Horsham. Anyone who knows Horsham will know there is a Railtrack engineer's yard and stabling for Thameslink trains beside the curve of the line to Dorking. A nice site for Horsham Nightingale Road, the terminus of my company's line. My first train is nearly complete, an 1850s era single plus five four wheel coaches of 1850s vintage. Quite suitable for a Horsham to Guildford via Dorking local. But devising the loco and carriage liveries has been fun