The problem is that its hard to schedule both passenger trains and freight trains on the same railways, and the benefits of transporting freight by train instead of truck are both economically and environmentally more important. OTOH high speed rails with their own tracks are awesome and let us get around this dilemma.
Perhaps I'm naive, but with today's technology (GPS, processing power for scheduling, communications, etc.) and a modest investment in some more switching tracks (compared to high speed rail) why couldn't we saturate the bus?
Because of geography and property rights this is pretty much a monopoly as a class of transport. And while the government's stewardship of rail thus far is not seen as a paragon of efficiency or customer service (http://reason.com/archives/2005/12/01/amtrak-sucks), the only way to make this work would be to have the tracks effectively nationalized.
The issue is that high speed passenger rail requires trainsets that are built basically like aircraft, and really do not do well in a collision with a freight train built and loaded like...a freight train.
It is definitely reasonable to use improved signaling to increase capacity, but rails for HSR need to be smoother, different grade limits, possibly banked, etc. They can often but not always share right of way with existing freight lines. Still, designing them to be grade separated, protected from incursions by malfunctioning freight, etc. is a hard problem.
I think the low hanging fruit is increasing investment in conventional freight, metro area rail for passengers, and limited investment in dedicated HSR only on routes where air travel faces capacity limits -- northeast corridor, los Angeles to las Vegas, and probably unfortunately not sf to la (as much as I would personally enjoy it).