>>1036876Philadelphia's subway-surface _tunnel_ literally handles 70-80 TPH during peak hours and allows 5 separate lines to neatly converge downtown without major hiccups. You couldn't replicate that on the surface no matter how hard you tried.
>>1036885Literally untrue. Karlsruhe, Cologne, Kassel, and Dusseldorf among others are all building Stadtbahn tunnels right now, and still expanding networks. They are simply smaller mid-sized cities with less emphasis on a central CBD like Melbourne's that everyone commutes to at peak hours, so they can get away with a more distributed system that only requires dedicated ROW where it matters. A 6-car 1000-person EMU is simply not needed or is already fulfilled by the S-Bahn, which like Metro Trains boasts the 40+ km regional coverage to actually fill up those trains. Stadtbahnen and light rail are mostly 20 km lines to the inner suburbs to allow for denser station coverage. Transit planning isn't simply about maximum capacity. Capacity is simply a function of how long a transit line is planned to be scaled. When regions are compact enough and planned frequency good enough, there are other options available at one's disposal.
While the single-line streetcar memes are a increasingly pervasive problem, I (
>>1028942) have already clarified their distinct origins from predecessor light rail networks like San Diego's, which are already reputable networks that are expanding as well. High-quality light rail like Seattle's will be improved to match the performance characteristics of Chicago's L, while published data on operating costs are also quite comparable to subways. Like any other fixed-guideway network, tunneling or building elevated light rail is primarily a matter of how much relieving constrained operation on the shared central segment will maintain frequency on the outer branches as a network grows. Such frequency and coverage is precisely what helps control the required vehicular capacity.
