The only line in the US that is anywhere near close to capacity is the Northeast Corridor, and Alon Levy wrote, "Everyone, don’t sell the NEC short. The line was operationally profitable until this year, when the recession killed ridership. If it were rebuilt at HSR speeds, it would make Amtrak as a whole profitable, creating a ready source of funding for further HSR construction."
Avi raised two concerns:
Alon, the NEC obviously has the greatest potential for ridership and profit. The problem is the cost to build a true HSR line would be astronomical. You’re talking about buying land and building a new right of way in the densest most expensive region in the country.
The other issue between NY and DC(especially NY and PHL) is there is too much traffic on the tracks. There is Amtrak Acela, Amtrak Regionals, Amtrak Keystones, random other Amtrak trains, NJ Transit express, NJ Transit Local, and Septa.
Avi clearly didn't read my earlier posts about redundancy, or he would know that there are parallel rights-of-way to the Northeast Corridor the entire way from New York to Boston. Some of them have commuter service, some run a few freight trains, but most of them are railbanked.
Eminent domain would only be necessary in cases where curves need to be straightened. The main expenses would be grade crossing elimination, electrification and rebuilding bridges. This is another way that redundancy can be useful: to run high-speed trains, you need dedicated express track. The Northeast Corridor has too many slower trains on it, but fortunately we have the right-of-way for another line.
So where could the Northeast High-Speed 2.0 Train run? From Boston south to New Haven, the best bet would probably be the Air Line, so named because it was almost as straight as an airplane route. The line is intact except for a bridge or two, and currently used as a rail-trail. From New Haven or Middletown the line could continue west on the former New York and New England line to Brewster and south on the Old Putnam line. It's not clear the best way to get it into Penn Station, but this is one of the many reasons why it's frustrating that the new Hudson River tunnel will not connect to Grand Central.
Between Newark and Philadelphia, the train can use the ex-Reading West Trenton Line. From Philadelphia to Baltimore I was a little worried at first, but looking more closely the CSX Philadelphia Division (ex-B&O) is single-track most of the way, and the NEC is only three tracks. There's plenty of room to add capacity. In any case, all of the alternatives are indirect, going through Lancaster, Harrisburg, York or Reading, PA, but some of the freight traffic could be shifted to those routes if necessary.
Going through Baltimore would probably be a headache, but that could potentially be done on the NEC. Between Baltimore and Washington we have three options, as I wrote before: the former Pennsylvania mainline currently used by the Northeast Corridor, the CSX Capitol subdivision used by MARC commuter trains, and the CSX Old Main and Metropolitan subdivisions, partly used by MARC.
The nice thing about these is that since we're serving the same cities, the alternatives could be phased in. I get the impression that the Shore Line is the most saturated section of the Northeast Corridor, but it's also the fastest, and the alternative Air Line is one of the most expensive to reconstruct. The most promising demonstration project would probably be the West Trenton Line. The right-of-way is already there, there is at least one set of tracks, there is some freight traffic but there are parallel routes for it to be diverted to. It also goes between New York and Philadelphia, where there are plenty of people ready to try it out.