Paul Druce is skeptical. He suggests 3.5 could be attainable, about the same as driving. I would be all for spending the money on a dedicated line, just to stop sharing the tracks with freight rail. We should fund it with value capture, obvs.
I’m not entirely certain that you could get under three hours for Pittsburgh-Harrisburg without spending a good deal of money; it would require an average speed above 80 miles per hour and while FEC is doing that, they’ve got a flat route that’ll be spending most of its time at 110-125mph in order to do it. Given the freight traffic on the line, you would likely need to build a dedicated high speed rail line in order to get it down below 3 hours.
However, Talgos and eliminating any low hanging slow zones might be able to get it down to 3.5 hours (Talgo is claimed to have calculated it at 4’10 without track improvements and retaining 34 minute padding). A second train frequency, preferably with an earlier NYC departure, is certainly called for however; occupancy is about 50% and there are definitely riders to be gained with an additional frequency.