Thought experiment...

What would it cost to convert mainline railroads to electric, and would it increase capacity any? What other things might be involved? With electrics, would it be reasonable to become more "European" with shorter, faster freight trains to fit better with expanded passenger service?

Several years ago, Caltrain estimated the cost of wiring the tracks (poles, wires, power supply-substations, etc.) at about $2 million/double track mile. That seems a tad low, but in the ballpark (say, an infield fly) for just the wiring job. Double that at today's prices.

Bigger issue for them is the rolling stock. Do you buy all new stuff (MUs probably) or do you just buy electric locomotives to replace the diesels and keep dragging the old cars? The latter's probably more cost-effective but you might give up a little ultimate operating capacity.

Equipment's a big cost consideration, but maybe not as much for freight motors. Nobody currently sells electric locomotives in the US (admit it: not much market outside Amtrak NE Corridor and that market is pretty much dead for a while). But diesels are just electrics that carry their own generators. So given that to a large extent an electric freight or even modern non-high-speed passenger loco would be basically a diesel with somewhat different controls, a bit of extra electrical gear (pans, transformers, etc.), and some ballast why should it cost (in reasonable quantity, not of course one-off) more than a diesel? Should cost less, and it should even be possible (if you really want to) to rebuild old diesel platforms to electric. For that matter, how about "hybrids" where an ultracapacitor bank is used to store energy from braking that then gets used for starting - you could do that with diesels too, but would have to find a place on the loco to put the caps while for electrics you could stack 'em up at the substation.

Power cost of course is a factor, and how many more (coal-fired probably?) power plants would be needed? For a passenger line, you have to run the trains when the people want to go, so you just have to suck it up and pay peak power rates. For freight, where you run a lot of stuff at night anyway, why not demand a break since you're using the base load plants instead of the peakers?

Suppose you have a RR where you can do most of your freight switching during the day with diesels (modern clean ones of course ;-) ). The main line during the day could be used mostly by electric passenger trains (letting the public pay the peak rates and contribute a disproportionate amount to the capital cost of the wiring of course). Then you do your mainline runs at night with electric power, possibly (for Z trains) using passenger power during off hours. Sound like some operations in the Bay Area or even LA?

So how much would you have to "wire" to make it worthwhile? 100 miles? 300? Let's see, 500 miles double track at $4mil per ... $2000mil ($2bil in American language). Sounds big to an ordinary mortal, but a $bil or 2 isn't really big for modern civil engineering projects of any scale. Add perhaps another $bil for assorted new/rebuilt lokies and odds/ends. That would cover, say, BNSF from Hobart to Needles or UP from LA to Yuma. Maybe even with Alameda Corridor thrown in. In the '70s SP did preliminary investigations for a Colton-Yuma electrification BTW - decided incremental second track and assorted mergers were a better deal, and of course diesel prices dropped like a rock in the later '80s which made 'em look smart for not electrifying. But maybe not so smart now...