In addition to what the cow said, keep in mind that other than one FTer's ability to check actual loads, we have little ability to access any of the underlying information that would truly help us understand the parameters that might make AC cancel a flight. EF and AC seatmaps are unreliable guides, as many people don't choose seats at booking. Inventory for sale is also imperfect; since AC is willing to oversell many flights by wide margins, even flights that are K9 may have a lot of passengers booked. It's true that a huge number of seats showing up as taken on the seat map, or very little inventory for sale, often indicate flights that are heavily booked, meaning that the data isn't worthless, but even in that case, there are many people leaving travel plans intact until relatively close to the time of the flight before changing/cancelling, so even if you see a flight that's J2 Y5 or something 3-4 weeks out, a rash of individuals cancelling their bookings could still lead AC to cancel what currently looks like a fairly full flight.

From an aircraft utilization perspective, many flights are out-and-back from a hub, meaning that the issue of where the aircraft is needed next may not be a big issue (it's going to end up back at the hub anyway, so if both flights have low loads, easy to cancel), but YOW-YVR is less likely one of those. And we have zero insight on cargo or on crew movement needs (which is not always the same as where the aircraft needs to be).

I had some huge changes/cancellations on bookings a couple of days ago just over 2 weeks out, and I've experienced a bunch of similarly late changes on my June/July bookings, so I wouldn't feel confident that they're done with August yet. Maybe once you hit two weeks, the risk might be low?

But who knows. Just brace yourself for a high likelihood of a change, then if your direct stays intact, it's a win