The problem is that population distribution means that almost nobody is going to be getting on or off the train between Minneapolis and Seattle. The population of North Dakota is 800k, South Dakota is 925k, Nebraska is 2 million, Montana is 1.1 million, Wyoming is 590k, Idaho is 2 million. That's nearly a whole quadrant of the country with less population than the Houston metro area. If we're building trains, let's build trains in Houston and serve the same number of people with like a tiny percentage of track that it would take to serve the upper plains states.

Why do I feel like inefficient access to the Pacific Northwest suits all involved?

And it's weird, because the map has some faint grey lines where you'd think there'd be routes.

From Japan rail networking, I've heard that you need a popular stop about every four hours for an HSR to be successful

LA is a bottleneck if you assume every single line and dot is perfectly equal. If we’re already imaging a well built system then that green line would have a higher frequency of train to accommodate what you’re talking about and it’s station(s) would be large enough to handle the fact that it would absolutely be a major hub.

Efficiency is not always about perfection for every single trip. Cars(in a car-centric hellhole, at least) will take you from your driveway to your destination parking lot but they are vastly inferior to the overall efficiency of a metro that you walk five minutes to and is then five minutes from your destination. This is highspeed rail, there’s not much extra time being taken if you don’t go direct direct, it’ll be fine.