You know, we really like to play around with grandiose studies, and then pretend that you can't do it because there isn't enough money.
Its like this:
"We need to buy a car."
"But a Rolls-Royce costs $350k and a Bentley costs $200k! We can only afford $50k! We can't afford to do this!"
"By god, you're right!"
Never mind that you could buy a mid-size Mercedes for $50k, or 2 full size loaded Ford Crown Victories, or 2 mid-sized Honda Accords, or 3 midsize Hyundai Sonatas, or 4 Kia Souls, or 5 Nissan Versas, or 10 10-year-old Mercedes e-classes, or 20 perfectly adequate, running used cars.
Nope. We can't afford a Rolls-Royce or Bentley, man. We simply can't afford a car.
I could get you running, for about $20 million, an Atlanta-Chattanooga train. It would use some old cars- ex-Metra bi-levels. It would use old but refurbished engines - ex-NJT GP40s or something like it. The track would be upgraded a little, but probably to no more than 50 mph average- but that's competitive with driving for that distance.
But then we have a train. And when funding becomes available, we can piggyback on an order for new cars- Bombardier bi-levels, or if Amtrak is ordering, Pacific Surfliner-style cars, with 125 mph trucks. Later, we can buy ourselves some decent engines. Alstom PL42s, if NJT ever places another order for those awesome items. Or less impressive engines, such as an MP40, if they don't. Later still, we can realign the track and try to get average speed to higher rates, 65-70, and blow the cars away.
And then we can say, over 15 years, it cost about $4 million a year in capital expense. And people won't sit there with smoke coming out of their ears about the silly $10 billion mag-lev.