If we're talking operational subsidy only then from memory pre covid Island Line had the highest subsidy at about 50p per mile. SWT/SWR was the only franchise to show a net positive.
Some things to bear in mind:
To get a true picture you need to measure subsidy to Network Rail as well as the operator. At privatisation the intention was that subsidy would go to the operator. The logic was that they were closest to the customer so best placed to spend it wisely. That's changed over the years with more going to NR.
It's very difficult to break down subsidy line by line as it's not always clear where costs and revenue should be allocated. For example what proportion of central costs should a particular line bear? In what proportion should a longer distance fare be allocated to different parts of the journey? And if you close a lightly used line do you then lose revenue for longer journeys as well? These arguments have been rehearsed many times over in relation to Beeching.
Finally it's generally accepted that road transport doesn't cover all its external costs, this means not just road maintenance and building but stuff like the cost of policing and healthcare for accidents. So while you could say it's cheaper to move everyone in taxis instead of rail it might only look that way because they're effectively subsided as well!
Although some of the subsidy to NR will be to do with enhancements to the rail infrastructure, some of which will lead to lower costs (such as electrification) whilst some others will lead to more capacity (both of which change the equation for future years).
I recently spent a few days in London and Uber was significantly less expensive than an old school minicab and a fraction of the "Black Cab" rate. I thought Uber was brilliant in my (limited mobilty) experience.
Uber is trying to tempt people from public transport, if that doesn't exist then process would be closer to the other modes.
I suspect that someone is subsidising the London Uber costs, as I doubt that minicabs are making significantly more profit (as the cars would be broadly the same cost to run) so it boils down to staff pay, level of profit, efficiency of cars (i.e. how much are they waiting about for) or outside funding.
You'd get few drivers if they are getting paid too much less than minicabs. Whilst Uber can make less profit per trip and will do very well overall it's unlikely to make a noticeable difference. Likewise Uber can make improvements to how many customers each driver gets, again it's not likely to be too significant. Now combined the above could make it noticeably cheaper. However it does appear that Uber are trying to focus on tempting people out of public transport by using outside funding.
The problem with any car based system is that you end up with almost all much congestion as you would have with privately owned cars (a bit less as there's a reduced need for on street parking).
Anyone that says that Uber or other tech based automated vehicle system will be able to replace public transport is probably exaggerating the benefits.
In reality there just isn't space for too many people to travel in their own little box when currently dozens or even hundreds of people are sharing a bus or a train. As the space inefficiencies would rule it out.
If you look at the Tesla system in Las Vegas and compare it to a tube train and you'd start to understand why it'll never compare in capacity.
At the weekend I joined a Jubilee line train at Waterloo, just in the doors I bordered at there was at least 8 people. The previous train was 3 minutes before the following 3 minutes later. Whilst it would be possible to have more vehicles than doors in that timeframe, the loading time of each car (especially given I was traveling with children) would be much higher (requiring a much much larger, and therefore much more expensive) station complex.
Tesla say they can reduce the tunneling costs (which they could do), however talk to anyone involved in Crossrail or HS2 and they'll tell you the big costs are not the tunnels it's the station boxes.
As such for lower level public transport (i.e. to get something working where there is nothing now) to compete with buses (especially if the buses aren't that week loaded) it may work, however if a full metro system is needed then governments are likely to just be wasting their money (which is ultimately our money from taxes).