The restrictions of rail scheduling, and the capacity constraints of Union Station, both in terms of pedestrian flows and rail capacity, won't allow for it.
If we're solely focussed in achieving the 5-minute timetable, the bottleneck will be rail capacity. It won't be pedestrian flows (post-revitalization).
For a 5-minute time table we're not talking about 12-car BiLevel GO trains anymore. It would need to be EMUs, the very thing Metrolinx said they will now procure for at least some of the fleet. I also think to achieve that, it likely would require making RER serve the UPX role (to use up the rail capacity that UPX is currently using up) where UPX ceases to exist as a "separate" monolithic service with its separate platform, but becomes rolled into RER like the Paris CDG airport spur of Paris's RER Line B and stops at the regular GO platforms. Costly as that may be, the decision of electrifying UPX may also automatically be part of deciding whether or not it is merged virtually into RER.
Many of the pieces are falling into place already to open the potential 5-minute timetable without the Union tunnel, but there would be some huge changes/compromises/etc that ends up being made to achieve that, like the bitter pill of discontinuing the short UPX trains/UPX stations during electrification, eliminating diesel trains from stopping at many stations, and merger of services, and a fast-accelerating EMU fleet.
Whether it should be done or not, is a good question.
But the pieces are already falling into places to
keep the 5-minute service option possible, should it be exercised.
- Metrolinx is very eager to roll out Positive Train Control as part of RER -- a system commonly found in many subways and European commuter systems -- but rarely in North American commuter systems. This operationally decrease headways more safely and easily. It is multiple sources (this, this, etc). If 5-minute becomes a political requirement, this step becomes a pre-requisite.
- Metrolinx confirmed this week they are purchasing EMUs (and livetweeted from this weeks' Metrolinx meeting). EMU is the same type of technology found in nearly all subways! They accelerate faster, and along with short-dwell, will help increase train throughput.
- Union Station Railway Corridor will get more train throughput after its resignalling.
- There literally becomes pedestrian expressways between Bay and York concourses post-revitalization. The massively wide TTC-Bay moat connection has almost two dozen doors (almost one dozen double-doors), practically twice as many as pre-revitalization -- there is doors along the whole length of Bay moat as a massive vomitorium (upper-right corner) between RBC-PATH-TTC-Union, and there will be plenty of space and multiple opportunities for passengers to divert stairlessly towards the right quickly towards York out of the crushing Bay flow (locating food in York also helps bait Bay pedestrians away from the crushing TTC-BayConcourse flow -- very clever). They intentionally also cleverly moved all coffee-lineups (commuter flow obstacle) away from TTC-Bay flows. And we all know that not all commuters will use the TTC-Bay connection, with all the additional PATH entrances planned.
- There are techniques to spread the instantaneous Union passenger bottlenecks around to fill the Union hourly capacity better (which will be much more massive). Many of these are already mentioned in the older Metrolinx 2031. People loiter too much in the wrong places, blocking flows. People surge to the stairs in the 10-minute timetable announcements. Shorter dwells and more frequent platform emptying with less-timetable-surging, combined with the currently closed-tracks reopened, to spread platform capacity around better while maintaining crowding to today's level.
There are many cost-plus options that will need to happen, and it might not fit in the SmartTrack budget. We might not be able to pull off 5-minute for more than one through route (e.g. only for Bramalea-Stoufville) except at peak where you can pre-peak pre-position 150-meter trains in double-berths before they head in opposite directions.
That way, you borrow USRC capacity in the previous hour to pre-load Union, so you don't have to worry about too many incoming trains during peak hour, when trying to maximize outgoing trains. LSW already does this to help maximize Union capacity -- you sometimes have an hour-long wait in Burlington to head into Toronto, if you arrive there just right before peak -- because they skipped that train to maximize Union peak capacity (in trainset count and to help reduce USRC contention issue). For 5-minute peak service, this might mean you'd get only offpeak frequencies (e.g. every 10-15 minutes) for incoming trains. At offpeak, double berthing would cease and the trains would continue in through-service (LSE-LSW and Bramalea-Stoufville). Today LSE and LSW already does this -- through service mainly occurs during offpeak.
The solutions that are applied towards Metrolinx 2031 will actually allow Union to have 2x the passenger throughput during peak. That's double the number of passengers walking in-and-out of Union during the peak hour!
Today's Lakeshore West runs approximately 7.5 minute headway during 4:45pm through 5:45pm, although they currently interleave into separate tracks, due to the long-dwell times of traditional GO trains.
TL;DR: Union passenger capacity (post-revitalization) is not the weakest link for 5-minute service. It's the USRC rail capacity -- how many trains can enter-exit Union.