To route a rail line towards Pearson, the costs would be pretty enormous but well worth it I think. You get a double benefit by being able to create a full HSR corridor that both Go and VIA trains could use. Connecting the Airport straight into the Georgetown line would have huge benefits, and the Airport's basically a necessity for good HSR service.
I think something interesting could be figured out with regional HSR stops. On top of the straight Toronto-Ottawa-Montreal route, there'd have to be a route serving smaller areas, like Toronto-Oshawa-Kingston-Ottawa. Maybe under that would be an even more local spacing, like Oshawa-Port Hope-Cobourg-Trenton-Belleville, etc. That would create a very strong rail network and get the best bang for your buck for HSR infrastructure. Express would compete mainly with the airline market that exists between Toronto Ottawa and Montreal, while regional and local would compete more with the highways and work to serve people in the smaller cities.
I agree that a on site station at Pearson would not be cheap, but well worth the money. Pearson is the most important and busiest airport in Canada by far and is hugely important to the GTA. To not create the most efficient HSR connection with Pearson would really be missing out on an opportunity to hugely improve the transportation network in the region. And I also think it should be fully accessible to GO as well, just as is the case with many other European airport stations (Schiphol and Frankfurt come to mind).
I think one good way of thinking about how to design the network is setting a limit on how much could be spent on it. I generally assume that the full Quebec-Windsor network would take 15 - 20 years to complete and that $1 billion/year would be an acceptable figure in the public eye to spend, giving a total cost of $15 - $20 billion. This is somewhat arbitrary but also based on general estimates that have been put out before.
So with that limit what are the most important aspects and where could trade offs in some areas be made to make improvements in others. One good example is the idea that line should go from Toronto to Ottawa and Montreal and completely skip a Lakeshore routing. Yes the total track length is less, but then you add an additional 70-80km to the total trip from Toronto to Montreal. And once you factor in slower speeds through some parts of Ottawa, this adds anywhere from 25 - 30 minutes to the total trip. Keeping a routing that is much like the existing VIA route means more money spent on infrastructure, but improves travel times between the two most important centres. And given that HSR would be built from scratch and should be about achieving the fastest possible travel times, it seems worthwhile to spend the money on what is really a small amount of extra infrastructure.
The same is true with making sure that smaller centers such as Belleville, Port Hope, Brockville, etc, retain their service and important nodes such as Pearson also have the most efficient connections. It will cost more money for passing track or new routings or spur lines, but it will also make the service available to more people. You can offset this with trade offs, such as not directly serving the Dorval terminal, offering some sort of connection for passengers that is quick and seamless, and leaving the Ottawa station where it is instead of spending massive sums on the tunnels that would be needed to serve the downtown area with an efficient station (and really to locate the station 3 or 4 kms closer).
And the good part about a well designed network is that service models can be easily changed allowing for as many or as little stops by as many or as little trains as is required. And if GO and AMT undertake electrification of even just their most important lines, that offers even more service possibilities to the HSR service providers.
One final aspect that I think is important but as far as I know has never been explored is consulting with cities such as Kingston, Brockville, Cornwall, Trois Riviere, etc to find out what they feel would be important from an HSR network and how local citizens feel about the idea. A lot of people see public consultation as getting in the way, but it is an inevitable step and working with them from the start should be done. Maybe some communities wouldn't mind a station on the outskirts with lots of parking, maybe some would prefer that local service be achieved by serving more central areas via a spur line. But no one really knows because its never really been discussed. Same is true of finding out how people in Richmond Hill or Laval feel about HSR and what they think would make the service useful. The answers would probably be interesting and revealing.
