There are a number of metrics where it's clearly better.
Scheduling and routing. Not waiting 10-30 minutes for the next train depending on time of day. Not stopping anywhere but your destination. No connections or changeovers.
Personal space. Privacy, safety. Regardless of how you personally feel about this factor, it matters enough to some people to significantly impact usage.
Those factors may or may not be enough to warrant changing the systems, but they certainly are metrics by which Musk's systems are better.
Waiting 10-30 minutes for a train isn't a technological consequence of rail transportation, it's a symptom of a catastrophically inept transit agency.
What's stopping, say, Caltrain, from running trains at peak frequency all day? Incompetent leadership. They have the trains (which spend all day sitting idle _at_ terminus platforms), the have the crew (who also sit idle during the day), they have the capacity to run more trains off peak. They made a deliberate choice to not run the trains.
You might make the case that running more trains off-peak would increase maintenance costs. But I would wager that such maintenance costs pale in comparison to the cost of building enough platforms at your downtown termini to store trains all day. Moreover who cares about maintenance costs? It's insane to think that public transit should be profitable.
> You might make the case that running more trains off-peak would increase maintenance costs.
Alon Levy has done the math [1] and found that providing extra off-peak service is about 20% the cost of providing extra peak service, since the capital costs and labor costs dominate the actual operational costs.
Today Caltrain can't run more trains because diesel locomotives are too expensive and slow, and the current route has no passing tracks which makes scheduling hard to work in with express trains, which are needed to stay competitive to commuters. When they move to more modern EMU trains over the next couple years, they will be able run with 10 minute headways, potentially down to 6 minutes if you include the blended service with the CAHSR trains.
But that only solves the waiting-for-a-train problem. Regional train line still have trouble keeping up with expressways because of the time needed slow down, wait, and then accelerate up to speed at each station. Lighter rolling stock can make this a little bit better, but not dramatically; BART has always used relatively light EMUs and it still can only average around 55 km/h. Even the most fanciful of Caltrain upgrades would have a top speed of 180 km/h, which means they'd really be lucky to average 100 km/h on a train making every stop between San Jose and San Fransisco. Express trains would be slightly better (perhaps using the HSR trains to do SJ<->SF nonstop in ~30 minutes) but that could only cover a subset of trips. For most trips, if you can avoid terrible traffic, driving will remain faster.
Well, they don't run them all day because of the cost of the use of the trains and the cost of the crews. The crews can't be paid all day to do nothing.
Besides, capital costs of construction and vehicle acquisition absolutely dominates the cost of operation. The problem isn’t that running constantly available service isn’t affordable, the problem is that we lack the political will to make mass transit a meaningful alternative to cars.
I only know about Seattle where they were not paying for no work in the middle of the day. They had a last 2 mile transit van that would take you from your house to the P&R and back for a while. It was because the park and ride was filling up earlier and earlier and so people couldn't get to the bus to ride it. In this case, I know because I asked them how the pay went. They only got paid in the morning and the afternoon, there was no service and no pay in the middle of the day, only during rush hour.
The drivers I spoke to about this didn't like this because there wasn't enough time in the middle of the day to get a different job so they ended up getting a longer than 12 hour day to get 2 4 hour work periods.
It’ll also transit a tiny fraction of what a traditional subway would, and requires ownership of a car to use (assuming that a Tesla isn’t a requirement or at least preferred).
I’m thoroughly convinced that these types of systems, along with self driving cars, are purely fantasies to allow people to believe that we can fix all the problems of car culture without having to give something up. If you want to move a lot of people through a city, build a subway. They’re proven and effective. Stop trying to find a way to turn subways into cars, the result is always ridiculous.
Or, given the capacity limitations of the boring company thing, just build a segregated bus lane. The main advantage of subways over trams and fully segregated bus lanes is capacity, so if you’re not doing that...
Is it really 2 minutes longer? Subways move millions of people a day. Your loop tunnel will have 1/100th of that capacity. There will be no traffic jams, waiting at the elevators, to merge / separate traffic? Are you addressing some fantasy scenario where you are the only person using a city-wide system?
Yep, except that in order to match the capacity of the Montreal metro system you would need to have a car departing every 2 seconds, which is impossible.
One car every 2 seconds sounds totally plausible. In fact, that number actually seems pretty low to me. A single lane/tunnel should be able to handle automated cars traveling at a 2-second following distance no problem, so that just leaves loading/unloading as a potential bottleneck. How many on-ramps do you think it would take to reach a capacity of 0.5 cars/second? I'm really curious as to what sort of problems you're envisioning which would make that "impossible".
You need people to drive their cars safely into the intake with an average delay of two seconds, to match a very middling subway system running at a delay of 7 minutes instead of 2, assuming that every single car is filled to the brim. This is completely ridiculous. You won't be able to make people do anything with an average delay of 2 seconds, without any mistakes. The question isn't on-ramps, it's getting people to merge or arrive at the dispatch on average every two seconds without making any mistakes.
Nobody will be driving in these tunnels. Only fully autonomous cars running approved software will be allowed in. Those sort of problems go from "impossible" to "almost trivial" when the exact movements of every vehicle in the tunnels is being monitored and controlled by a centralized system.
They still need to get somewhere in order to enter the system, unless you're banking into full self driving. And they need to get there at a pace of one car every two seconds, or there needs to be a buffer.
How many amusement parks can load a ride car once every two seconds? Even continuously moving ski lifts aren't anywhere close to that and that's basically ideal conditions where the seat is delivered straight to your butt, and most users can reasonable be expected to be fit and coordinated. In the real world, people dawdle. You'd probably need dozens of loading lanes to reach .5/s, which means the stations are very large and expensive caverns.
Loop is basically a personal rapid transit system, which is an idea that keeps getting proposed for basically the reasons you cite, but the few implementations that do exist have generally underwhelmed.
https://twitter.com/elonmusk/status/1285819565407002624