If you have a wholly and exclusively public sector railway, run as a public service in the way the police, or your local council is run, then ROSCOs are pointless. However, this would be unusual. In the UK even functions such as schools, the NHS, Highways England, etc are not run in that way.
Your local High School, for example, doesn't just employ teachers and send the bill to the government. It has a budget, and it has to allocate spending between capital spending and day to day spending, just like a company does. Same with NHS trusts, etc.
In simple accountancy terms, it makes no sense to say - for example - a school costs £4 million to run in 2024 and £300,000 to run in 2025, when the 2024 figure includes the cost of buying a new sports hall and fixing the roof. Firstly, that sports hall and that roof will last a long time and secondly the school wouldn't function without having its buildings. So in reality the costs on things like land and buildings are not just part of the costs during the year when they are bought and built - they are costs that apply throughout the lifetime of the assets. Hence the difference between capital expenditure and revenue expenditure.
So your capital expenditure is spread across a number of years (even though you actually spend it only during the first year) for accounting purposes.
If you are a TOC running on a business basis - whether or not you are state owned or private sector - the way this is done affects your profitability. In particular for private sector TOCs it also affects your tax, because the current tax code does not allow you to offset the capital spending against tax. Instead you are only entitled to what are called "capital allowances". So if you spend £10 million on something that lasts 10 years, and make £1.5 million in revenue each of those ten years, a common sense view would say you'd be able to split your £10 million costs over the 10 years and pay tax only on the £500,000 difference. But you can't. You can only claim a certain percentage depending on very complicated rules and then make a final adjustment when the asset is sold or scrapped.
On the other hand if you lease an asset for £1 million a year that is counted as expenditure and you only pay tax on the £500,000. So unsurprisingly leasing is very common in the business world for everything from photocopiers to vans to offices to aircraft. There are tradeoffs of course, but a system with private sector TOCs is going to have leasing companies. A system with public sector TOCs is probably going to have leasing companies, both because they prepare accounts in the same way as a private sector company would and also because the debt incurred in buying the asset in the first place does not count as public sector borrowing. It is not like buying or renting a house in a rising market (like the UK) where a guy at the pub says "you should buy a house not rent it otherwise it's dead money".
The next issue is "gearing". If your business (whether public sector or private sector) produces £10 of benefit for every £5 you put in, if you put in £5 million, you'd get £10 million out. If, however, you had to spend £4 million of the £10 million on buying something, you'd only get £2 million out. So you don't. You borrow the £4 million to buy your asset and get £10 million out each year, less the cost of borrowing (interest) and capital repayment. Every pound spent upfront on buying things that you could actually borrow and pay back over time is a pound you aren't spending on something useful today.
So that leads to the next issue. Who can borrow the money more cheaply and efficiently? If you're the USSR, Venezuala, Zimbabwe or the Liverpool City Region, then the answer is the government. But in reality some government debt is linked to interest rates and some to inflation both of which have been extremely high as most people know. Consequently the government is paying interest over 4% on national debt in the last year. The more the government borrows, the higher the interest gets, and current government debt is over 100% of the entire economic output of the country and the highest it's been since the 1960s (when we were paying off the costs of World War II). The government still borrows £85 billion a year more than its budget - and due to high inflation and interest-linked bonds last year had to pay £112 billion on interest. The only saving grace is that due to Covid every government is in a similar position, so we don't really stand out - if we did it would be begging bowl to the IMF to rescue us time.
As we aren't the USSR, even if the government borrows the money, someone's getting interest on it and making a profit out of the deal. There is a cost involved in administration and the like. And due to the pressures referred to above, there is every likelihood that the government would say "OK we will fund purchasing of 1,000 trains, but we aren't paying to replace the local hospital in XYZ town this year". It's not a magic money tree. Indeed the most likely answer, as mentioned up-thread from BR years is "No, we aren't funding 1,000 trains, you can have 150 and go and buy some old Leyland National bus bodies and stick train wheels on them if you want anything more".
A private sector lender can borrow what it can, without any of those considerations. It can also sell and re-sell and leverage its future income stream, buy futures, take gambles on exchange rates, do deals with manufacturers, etc., etc. A private sector lender also can go bankrupt altogether and its creditors be left with pennies on the pound and shareholders left with nothing. Railtrack PLC tells us that the railway is not immune from that.
And sometimes, no doubt linked to the issues relating to interest rates and inflation, the private sector lender can make a huge profit as well.
But in all honesty so what? If they provide the service required of them, and its cheaper than borrowing by the state (as Transport for Wales discovered), and it isn't subject to government pressure due to high public sector debt, what is the issue? You choose your poison.
Do you want to get a set number of trains provided at a set price that you can plan for over a set period of time and which is tax efficient and / or allows a reliable prediction of the cost of the service? But someone, somewhere might make too much money off it for your liking?
Or do you want to beg the government to prioritise your plans for trains over a thousand other priorities and risk having them come back in three years time and decide it's too expensive and transfer the trains elsewhere and it be possibly more expensive and not get approved as desired (if at all)?
They're the options. The option of "the government gives us everything we want, but because evil ROSCO barons aren't making a profit all drivers and guards get an immediate 70% payrise" does not exist.