That being said, I think even people that agree with that statement would still be surprised at the kinds of stops that successful systems make. For example, Tokaido Shinkansen, the most financially successful line in the world, makes stops in Fuji (pop. 245k), Odawara(195k), Anjo (188k), Kakegawa (117k), Nakamura-Ku (135k), Mishima (110k), Hashima (68k), Maibara (38k), and Atami (37k). There are three services: all-stop, skip-stop, and non-stop. The fastest non-stop service is about an 1:40 faster than the all-stop service, and 1:00 faster than the skip stop service. But most terminal-to-terminal customers still take the skip stop service because it has the highest frequency...enough to negate the speed advantage of non-stop service in most cases.
When the CAHSR proposal first came out, the phrase "cow towns like Merced" was practically a meme. People were laughing at the stop in Gilroy, saying things like "why? so we can buy garlic on our way to LA?". Most people even thought that stops in secondary cities like Fresno and Bakersfield were entirely unnecessary. But the experts who wrote the report (who are fairly well regarded, btw) chose those stops for a reason: they are rural minihubs that aggregate a lot of traffic volume in areas that are too far from larger cities.
In the words of an operations researcher I once worked with: with HSR, top speed is what sells the service, but acceleration is what makes the money. Acceleration minimizes the time loss from making more stops, and more stops means more money.
Trains have a number of advantages, one of which is a small time penalty for stopping at a station relative to aircraft (2-3 minutes vs ~45+ minutes on average), and the aircraft requires a lot personnel/all-hands to turn around quickly (see: Southwest Airlines). A commercial aircraft is simply not designed for lots of really short hops in rapid succession. There needs to be a new preflight check, refueling, unload/load belly cargo, some maintenance items are based on aircraft takeoff/landing cycles (tires, cabin pressurization, etc.) Trains stopping at station and continuing is trivial by comparison.
A rail operator wants to collect fare paying passengers along the route they are planning on operating, it’s an easy way to boost ridership and revenue. It is a balancing act to have station stops, but not increase end-to-end trip time excessively. Basic rule of thumb: Does stopping add more riders than it costs due to the stopping time penalty? If the model/data say yes, make the stop; otherwise do not.
And just because we have a station, doesn’t mean every train has stop there. We can have local, limited, and express trains. We can passing/siding tracks at stations to prevent local stopped trains from blocking the mainline for limited/express trains. We can have cross-platform time transfers between different train types and destinations. We can have multiple mainline tracks. These things do cost a bit more upfront, but it’s more about good planning and what kind of service to the public is the goal.
— The Tokaido corridor (Tokyo-Nagoya-Osaka) is extremely dense, both the Tokaido Mainline and Shinkansen line are effectively at capacity (which is also part of impetus for the Chuo Shinkansen). Running at very high speeds with fews stops, consumes a lot track capacity. Which is why JRs prefer running more local and limited-stop service trains, and less than “non-stop” express trains. It’s what a capacity constrained operator should do to optimize operations and maximize capacity subject to those constraints.
Likewise a California HSR program, should find a way to cost effectively serve cities like Fresno, Bakersfield, Merced, etc. as part of a larger HSR network.
Too bad I don't work in that office anymore.
Within Tokyo most of the trains run every 5 minutes or so.
Unfortunately this system means you need parallel tracks, one set for the high speed express trains and another for the regionals. This isn't as bad as it seems, it requires double the steel and concrete, but the land is the expensive part of the deal and even though the parcel you need is wider the expense is not doubled.
Or you can run the high speed rail on routes that are optimal for it and leave the regional rail in its byzantine routes from the 30s that serve the locals well.
Why is that? Cars and airplanes accelerate much faster, without any real rider discomfort. If it's due to train hardware itself, wouldn't an assist to get the train up to speed work, in a fraction of that distance?
I just redid the math - to get to 300 km/h needs 4.8km, not 10 that I thought. Faster speeds need more distance or course. Hills also impact the distance needed. You also need to account for same safety distance for a lot of acceleration applications.
How much time is added to the journey for a single stop (including time for the stop itself, braking and acceleration).
That said, the advantage of rail is it stops at points in between. It cannot stop at every little town, but it needs to stop at every big city (you might even be able to do a suburb to downtown - the idea looks like it would work on paper but I'm not aware of real experience), and it the big cities are far enough in-between some small city should get a station.
This is very complex though. There is a trade off between stopping time and number of people on the train. You can stop in each city for as little as 1 minute (plus a few more to decelerate/accelerate) if your train has a lot of doors making it fast/easy to get on/off, but those doors mean there are less seats so less people are on the train. Or you can have just one door, but now there is a long line at that door and you need to stop for 15 minutes at each small town. So busy lines will have less stops than not so busy lines - the latter need more stops to attract all the customers they can get.
This trade off needs to be considered separately for each line.