I think the BIGGEST part most people are missing is the passenger + freight option. At a slightly higher estimated build cost, this option presents a massive increase to demand that could utilize the off-peak passenger travel hours to further increase revenue. Under his proposed rough calculations he determines the ticket cost by spreading the total estimated costs over 20 years. But if you can increase revenue you could reduce the timeframe significantly. This proposal is also only focused on a single closed loop system. Now imagine several stations and interconnecting tubes. More routing options will lead to greater usage. This will also lead to a reduction of Interstate traffic. With less interstate traffic, existing lanes could be utilized for additional tubes between existing stations. Also, since we're building above the interstate, why not double the pylons and build a double-decker tube system? The point is that there are multiple options to providing increased capacity. Additionally, any real station should be built with the potential for multiple connections, meaning incoming and outgoing vehicles would be routed based on demand and efficiency. To do that you would have more of an airport type of design with loading and unloading terminals away from the main transport section.
The point is, raw maximum capacity isn't the only consideration. In fact, if you look at every other modern transportation mechanism you'll see that maximum capacity is rarely the most important factor. Profit can often be achieved with sub-optimal processes.
I actually have a lot of more detailed ideas on how such a system could work. But most people on here wouldn't read past the first sentence.