We're looking at a (theoretical) $8 Billion system... that is probably grossly underestimating costs at that... that can only support 1120 passengers at a time, max. On the entire system.
Hyperloop may be fast, but it transports very very few people. A single-train will transport more people. Frankly, when looking at mass-transit, the #1 concern is moving people, lots of people and quickly.
Just like people today care about the "bandwidth" of their internet connections, I guarantee you that city planners are more interested in the "bandwidth" of a high-speed system.
At an 80min round trip running 20hr a day that's 15 trips per pod per day. So we have 40x28x15 = 16,800 daily round trips.
The Acela holds 360 people per train with 16 round trips per day. That's 360x16 = 5,760 daily round trips.
A high-speed system with dedicated track could run more often, and you could use longer trains than the Acela does, so the potential maximum bandwidth of high-speed rail is higher. But 16,800 daily round trips is large enough to be taken seriously.
(Why do I say "if it were possible"? http://pedestrianobservations.wordpress.com/2013/08/13/loopy... )
Really, Hyperloop needs to be compared against the entire Northeast Corridor... a single rail-line between Washington DC and Boston... but this single-rail serves many many more trains than just the Acela.
Or at very least, the comparison needs to be against Northeast Regional, a much slower train but is the one that your typical consumer is going to take. After all, Acela trains share its tracks with Northeast Regional... and most people prefer the cheaper prices.
(Between Boston and NYC, the link I'd most likely take, the Acela is only marginally faster than the Regional. The improvements planned in CT look like they would help, but actually cut similar amounts of time off the two routes.)
IIRC, Northeast Regional is considered to be at capacity around NYC (where it services on the order of 100,000+ daily commutes).
Anyway, the reason why this is being compared to railroads, is that the investment here is not measured only in dollars. When you build ground-based transportation of any kind, it necessitates blowing over some people's property... period. That is a real cost in people's livelihood, as the government forces people to move out of their homes to make room for the new road / railroad / hyperloop.
So when we're talking about an entire "Hyperloop" system that can only service ~16k / day in theory (and probably much much less in practice), these numbers have to be weighed against the costs.
True, both Railroads and Hyperloop suffer from this problem. But I bet you that the proposed California rail system will have significantly higher capacity... especially when the Rail system is going to connect up to the local metro systems.
In fact, the High-speed rail system goes directly to Union Station in the heart of LA. Hyperloop doesn't have such plans.
-----------------
Airplanes are magic stuff though, because they don't have these costs associated with them, and in practice Rail isn't much more efficient than Airplanes anyway. But the airways are crowding up between the cities, so building alternative means of travel is necessary now.
When you have two garbage estimates, you cannot compare the two systems. The solution is to ignore the bad estimates, and recalculate the costs more realistically.
The fact that Hyperloop is modeled as a three-pylon design, and totally ignores some basic physics (such as Thermal Expansion: http://www.leancrew.com/all-this/2013/08/hyperloop/). So it is clear that the Hyperloop cost numbers are just straight up irrelevant garbage.
Granted, that is to be expected from an unbuilt, untested system. So I don't want to hate on Musk on this one. The nature of proposing an experimental design is that you won't have good data on it: costs, problems, theory... etc. etc. No one really knows how Hyperloop will _really_ work.
The benefit of High Speed Rail is that the systems have been built around the world. They have a better grasp of how much it is going to cost by nature of building their systems in other locations... and having experience with the problem.