This is hard for me to do because 1) I probably respect Musk more than any living engineer, 2) I can't stand the CHSRA and their incompetence, and 3) I'm arguing from my comparatively unsuccessful keyboard against someone who is in every way my superior when it comes to engineering and business, and I know I'm gonna be the one to look like a fool here. But I'm going to do it anyway and try as best as I can to avoid the middlebrow dismissal.
Hyperloop is a nerd dream and it doesn't have a chance at financial success. The fundamental problem has nothing to do with the technology, which appears amazing and viable. If anyone can overcome the few nagging engineering issues, Musk and Co are the ones to do it. No, it has to do with not having a viable market plan, suffering from very common delusional assumptions in business models that have fucked over railroad capitalists for well over 2 centuries now, including three major railroad bubbles (Railway Mania as well as the Panics of 1893 and 1873). [1]
The faulty assumption is that the viability of the railroad hinges on the ability to go from major city to major city as fast as possible. This is only partially true. Of course you want to be as fast as possible, but the worst possible way to be as fast as possible is to eliminate stops in smaller cities and towns, which is exactly what Hyperloop is planning on doing. Those small towns usually end up being the most prominent drivers of revenue, overshadowing the large-city-to-large-city revenue by an order of magnitude. Even the CHSRA business plan shows that their revenue model does not depend heavily on the SFLA passenger volume.
You can typically think of demand for travel in terms of cultural dependencies. I have a need for employment, therefore I travel to work. I have a need to visit family, therefore I travel to family. I have a need for tuna, therefore I travel to the wharf market. I have a need for a visa, therefore I travel to the embassy. These cultural dependencies result in a demand distribution for travel that roughly fits a power law profile. The demand for near travel (x > 0) is exponentially greater than further travel (y > x). It is not hard to speculate that the passenger miles travelled by Angelenos traveling to work outweighs the aggregate passenger miles travelled between Los Angeles to San Francisco on any given day...LA County's Metropolitan Transit Authority alone accounts for 5.4M passenger miles a day [2], approximately equivalent to 14,000 passengers traveling between SF and LA. More importantly, the larger the city, the more your cultural needs are met by your own city. If you live in Los Angeles, your probabilistic need to travel to San Francisco is much lower than someone from Fresno, because your needs are mostly met by Los Angeles. When I lived in Stockton, I travelled to Oakland or San Francisco almost on a weekly basis...but now that I live in Seattle, I travel outside of the Puget Sound region maybe 2-3 times per year.
In the very beginning, this faulty assumption was actually a safer assumption to make than it is today, due to the lack of airlines and automobiles. The railroad literally was the fastest way to get from city to city, with no exceptions. But now it isn't. Once you go over about 300 miles, a well timed trip by plane is often as fast or faster (a lot of people will say that 500 miles is the threshold of competitiveness, but travel market share peaks and then drops off after about 300 miles). Under 100 miles, the freedom of a car to leave without a schedule often makes cars faster. Any time you have competition, you lower your market share, with your comparative competitiveness determining how much market share you lose.
So this is how it has played out hundreds if not thousands of times in the history of railroads: 1) Some big city businessman sees a slow trip to a city that he commonly has to travel to. He sees a demand for faster transportation. He throws together a business plan that makes the correct assumption that City A has X people, City B has Y people, and a fast connection between them will yield Z% market share of the travel D demand between them. He figures that in order to get that fast travel time, he can only stop at the largest cities in between...or possibly no stops at all! He then looks up a similar city or railroad (lets say B & S railroad), looks at ridership figures, and tries to back into D, and then inflates the Z figure to reflect the fact that his railroad will be faster because there are fewer stops. Therefore his revenue model becomes X * 0.5D * Z * P + Y * 0.5D * Z * P (where P == ticket price) 2) He sells some stocks and bonds (or in the case of modern capitalism, lobbies for grants), and builds a railroad. From this article, it looks like this step is starting. 3) After a few months, he realizes his model is inaccurate. At this point, he will do one of two things: either advertise more or remove stops to make it faster, both of which exacerbate the cash flow problem. 4) After about a year, he is bankrupt. The bondholders reclaim as much as they can by selling the railroad to some other capitalist who will with high assurance do exactly the same thing, but with a greater chance of success due to a much lower investment. And thus plays out the railroad consolidation game.
In terms of the model, his failure was actually three failures multiplied together. 1) He assumed that the ridership of B & S railroad was primarily between termini. This meant his D figure was too low because he vastly underestimated the demand to and from the small intermediate cities. 2) He assumed that B & S ridership figures were rides at full terminus-to-terminus prices, therefore inflating P. 3) He assumed that D was a linear function of population, ignoring that the larger the city the lower the cultural need for travel, thereby inflating D. The only thing he got right was X and Y.
The practical implication of this is that if you want a successful, profit-maximizing railroad, you need to optimize the need for speed with the need to pick up passengers along the way. If you are a capitalist that cares about maximizing profit, this unfortunately means having a railroad that looks like a slow kludge, because it means making unglamourous stops in unglamourous towns. And it also means that technophile idealism puts you at a competitive disadvantage.
[1] for more information/history, this is a great writeup, with a couple chapters that clearly describe this delusion. It is very long, but amazingly interesting and worth reading outside of this reference for anyone interested in technocapitalism or bubble economics. http://www.dtc.umn.edu/~odlyzko/doc/hallucinations.pdf)
[2] http://www.apta.com/resources/statistics/Documents/FactBook/...