If we have any intention of connecting the Canadian and American passenger networks across the Ontario/Michigan border, then a customs station at Michigan Central is (from what I can tell) the only reasonable path forward. Of course it would be more ideal for Chicago-Toronto passengers to have a customs facility in Toronto, with a single-seat ride seamlessly crossing the border, but doing so would also require customs facilities at any Canadian station at which the train stops. That is a good longer term goal, but even then the first step would still be to build a customs facility in Detroit to serve all of the communities between Toronto and Detroit skipped by Amtrak's through service. If we only build a customs facility in Toronto, then everyone in southwestern Ontario would need to backtrack to Toronto to cross the border, which of course hardly anyone would do. The number of passengers lost west of Toronto could easily exceed the number of passengers to/from Toronto who are dissuaded by the travel times at the border.
The key factor which makes the Detroit hub concept more palatable than some other en-route transfers is that most passengers would want to disembark in Detroit regardless, either to go to/from Detroit, or to transfer to another service such as the existing train service to Pontiac, the planned commuter service to Ann Arbor, and the proposed Amtrak service to Cleveland via Toledo.
Conceptual services from Michigan Central station.
- I transfered the existing Pontiac Amtrak service to the future Michigan commuter rail agency, to allow through-running from Ann Arbor to Pontiac serving both of Detroit's central stations.
View attachment 376487
The primary goal stated in the above article is not to create a Chicago-Toronto rail connection, it is to create a Detroit-Toronto rail connection, with the through-traffic from other American cities being a modest bonus. The US is already planning to increase service between Chicago and Detroit, and putting Detroit on the VIA map would allow them to increase Detroit/Windsor - Toronto service too, regardless of any through passengers from Chicago.
A service with connections in Detroit will certainly not be very fast, but neither are the alternatives. Wait times at the roadway bridges are often long and unpredictable, and plane passengers also need line up to clear customs. Both of those delays would tend to be well under the 90 minutes likely to be scheduled for a transfer in Detroit, but much of that time could be spent wandering around the new amenities at Michigan Central, which will include shopping and parks.
Indeed Canada needs to be fully on board to make this happen, primarily to upgrade the Essex Terminal trackage for passenger use. However, the US could fund most or even all of the capital work to reactivate Michigan Central Station and build a customs facility.
If we do decide to go through with upgrading the Essex Terminal trackage, we should indeed aim for a lot more than just a single connection per day. Perhaps one train per day could run non-stop from Detroit to London to avoid the reverse move, but other trains should also be extended across the border.
Here's a vision of how a relatively modest service level on both sides of the border could massively increase cross-border connectivity. The idea is that every 3 hours there would be a timed meet, where all or most of the lines would arrive and dwell, facilitating transfers. To allow time to clear customs, the VIA train would be first to arrive, and last to leave.
I assumed 4 platforms in service, because Ford has stated that they are protecting the 4 southern platforms for passenger rail operations. This is rather underwhelming - I would have wanted at least 5 platforms to enable growth for the 5 services serving the station.
View attachment 376511
The below timetables are purely to illustrate the concept for timed meets, and the times of day at which they might occur. Any of these time slots may actually be occupied by other train movements and/or not be approved by the host railway. Travel times between Detroit and Windsor are also significantly reduced compared to current trackage.
I made these fantasy timetables a while ago and am not entirely satisfied with them, the dwell time for VIA should be much higher to provide a larger margin of error for passengers transferring through customs.
View attachment 376503
View attachment 376506
View attachment 376505
The station access in Detroit is certainly an important consideration. In addition to looking into options such as extending the streetcar along Michigan Ave to the station, it's worth noting that people also have the option of taking a taxi/ridershare to the station. The cost of those services makes them impractical for access to regional stations (*cough* London at 05:20 *cough*) but even Americans are accustomed to travelling to a car-centric wasteland (a.k.a. airport) without using their own personal vehicles, as part of a long-distance trip.
Given that so many passengers will be embarking/disembarking at Detroit regardless of how the customs is set up, I would argue that Detroit is actually is a more promising place to demonstrate an international rail terminal than Niagara Falls. Niagara could then use
Detroit as a model.