OK, power/weight
is important for these motors. I was thinking the separate locomotive engine needs weight to get enough traction to pull the entire train. Like this hybrid yard locomotive that had to add ballast since the 50,000 lbs of batteries were not heavy enough!
http://www.railpower.com/dl/GGSeries.pdf
However, with distributed traction there is no separate "engine," and reducing motor weight provides more payload in the cars (since the individual cars each have motors). Since this weight is unstrung, it is even more important to minimize it.
Besides the peak vs. continuous ratings, another reason for the lower kW/kg spec is likely the durability requirements for locomotive use. Automobiles will be scrapped after 100,000-300,000 miles of use, with an easy duty cycle that uses high power every once in a while (sometimes never), and only for short periods. Locomotives likely are expected to run 1,000,000 miles or more (typical requirement for a line haul truck), and have a much more severe duty cycle, running at high power for several hours, every day. This requires heaver construction.
This durability/duty cycle difference is why a 400 hp diesel engine for an 80,000 lb. truck is much bigger and heaver than a 400 hp diesel engine for a 10,000 lb. pickup truck. Same goes for their transmissions and rear axles.
I think this helps explain why the TGV/AGV kW/kg is much lower than Tesla's, even thought they are trying just as hard and are at least equally competent motor designers.
I'm sorry if this is off-topic. Maybe we could rename this thread "planes, trains, and automobiles..."
GSP