The Interstate II proposal looks like something that would want to alleviate the problem not off commuting, but of short hop travels. The problem is getting these stations into the downtowns of major metro areas. A simple subway in SF to the Transbay Terminal would double the budget of any project like this due to the cost of real estate. I'm sure it would have similar effects in other metro areas. The grade separations would also create problems as city centers are already broken up by the city, and as we learned when we built the interstate highways, the people who were displaced because of construction were the people who literally had no place to go.
For something like this be effective, it needs to be done in a few prime corridors to prove it can be successful. For instance a highspeed line connecting Vegas and LA. Once people see it can work, people will use it. Building a nationwide system that is unproven in American terms would truly be a boondoogle.
But these stations have been in the major city downtowns for over a century. The CalTrain terminal (formerly SP) at 4th & Townsend is an example. So is the Oakland depot at Jack London Square.
I've always liked the idea of something like this LA to Vegas idea. This would work because between LA and Vegas are only maybe 3 stops (somewhere near San Bernardino, Victorville, and Barstow). But one thing I haven't seen mentioned is setting it up so you could load your car onto the train cheaper than you can drive. Without stopping much you could do the trip in about 2 hours and have your own car on either end. I think then it would work. It might even be faster than flying, when you count the checkin and checkout.
I think that this is where the article goes wrong. He mentions intermodal, but doesn't make it clear about what he is intermodal with. I think he is thinking airports and buses. Why not cars/interstates also?