Then I sat at an intersection one day and counted how many individuals were being moved through two lanes of regular car traffic vs how many came through on the light rail.
I was shocked by how inefficient regular traffic is.
So now MY gut feel is that without any technological breakthrough, this sort of system would already be starting with an order of magnitude disadvantage: which is to say, shockingly inefficient compared to allready available technology and solutions...
People are walking from the station to these giant office towers in central London. The walk from the Tube station to the office must only take 5-10 minutes, total. My gut feeling says the throughput won't be high specifically because you'll have 800 or so people trying to get out at the same place at roughly the same time.
Where mass transit struggles is in getting people from one arbitrary spot to another at an arbitrary time.
There is a delay between each car which after 20 cars really adds up.
How would that work? Now it's true that there won't be any delays caused by drivers not paying attention, but unless the cars will go 125 mph with only a couple meters of space between them (eep!) the cars will require a non-zero delay "off the line" to allow them to spread out as they gain speed.
How fast they can fork and merge is more interesting, and I dare not have an opinion about that.
Another problem is that all those people have to walk to and from the end station.
A "packet switched" system of sending individual cars at high speed solves both those problems, and going just by my gut feel, should have a quite high throughput too.