+ My opinion about the utility of compacts is that there is no evidence on the ground that they will lead to interstate high speed rail networks. Beyond evidence on the ground, the political requirements for interstate high speed rail is sustained political support for the network in each state in the network. Assigning a probability (p) of sustained political support to each state (S) in an interstate network (IN), we have:
pIN = pS(1) x pS(2)...pS(n)
I consider
intrastate high speed rail an optimistic proxy for sustained political support for
interstate high speed rail. In general, that means pS(i) approximates zero for all 1 intrastate high speed rail obtain.
+ The Great Lakes Compact is legally binding. It required approval by each state's legislature. To me, the politics are radically different. The compact prevents states from damage by other states. It exists to avoid a tragedy of the commons where people may die. Rail networks don't have that.
+ I considered diving into the differences between the 1940's and today as they relate to passenger rail networks in my previous comment.
Local passenger rail isn't viable because most rural communities have much lower populations today than in the first half of the 20th century and because transportation alternatives are much better in rural areas. Hopefully, we agree on the second point.
The demographic shift in rural areas has a strong labor component related to mechanization of agriculture. In the midwest, a farm family in 1920 could work about a quarter section of corn (160 acres). Today a family can work about two sections (1280 acres). A small town rail station in 1930 might have served a township with 3000 residents. Now it might serve 1000.
In the South's cotton country, the demographics are even more substantial. In 1940 a share cropper family could work about 20 acres. Mechanization moved major cotton production to large fields in places like Texas. The twenty acre fields were mostly converted to timber and the share cropper shacks left to decay. Small towns that once were the center for several thousand locals are now empty buildings with a few hundred locals nearby.
Local rail in rural areas is economically untenable. The express trains of 1950 worked because there were finer grained networks feeding it. Today, the fine grained network for rail is the same as for air and automobile. It's the automobile.
That's why intrastate high speed rail isn't being built. The proposed Orlando-Tampa route still required a car to get from Azalea Park to the Orlando station and another car to get from the Tampa station to Clearwater Beach.
+ One of the major tradeoffs between rail and auto transportation is scheduling. Scheduling trains efficiently is a job-shop problem and is NP-hard. Adding efficient public transit at each end of the Orlando-Tampa line to avoid the need for an automobile adds terms to an already exponential scheduling problem.
The way to mitigate latency in scheduling problems is not cleverness. It's adding substantial surplus capacity to smooth out peaks. It's making the system generally inefficient. Amtrack's poor level of service is a natural result of focusing on efficiency. High speed rail systems designed around efficiency will not beat the mathematics. Nobody is making a high speed rail case around high availability because it's politically infeasible.
It's probably economically infeasible as well. Rail is a circuit switched network. If it is designed to handle 1000 passengers per hour, handling 1001 passengers takes two hours not 1.001 hours.