First of all, any rail line which hasn't seen passenger rail service for almost 40 years* will require substantial capital expenditure to upgrade the infrastructure to a state which may allow the safe operation of passenger trains at travel speeds which allow travel times which are remotely competitive to driving and there is no reason to believe that CN will invest into a line which is of only tertiary importance within its network, just because VIA decides it wants to move its passenger trains onto that line. Even worse, the Southern half of the Saskatoon-Regina line (i.e. south of Davidson)
appears to be owned by a Short Line (
Last Mountain Railway), which usually indicates even worse track conditions and less eagerness (or rather: financial capability) of the host railroad to invest into these tracks...
*The last passenger train operated between Regina and Saskatoon in 1984, when the Winnipeg-Regina-Saskatoon day train was replaced by the “Panorama”, an overnight service which operated daily between Winnipeg and Edmonton on the route of today's Canadian (i.e. via Melville rather than Regina) and was extended three times per week via Jasper and Prince George to Prince Rupert (refer to the October 1983 and June 1984 schedules in my
VIA timetable archive).
***
Concerning your actual question, track access charges and maintenance costs are directly dependent on train mileage, whereas a longer route usually also results in a longer travel time, which increases labour costs, especially if longer shift times increase the risk that Locomotive Engineers run out of hours and need to get relieved.
To estimate the impact on costs, let's just assume that all operating costs are proportional to distance. Thanks to
VIA Rail's Summary of the 2019-2023 Corporate Plan, we know that the Canadian's direct costs were
$66.7 million in 2018. By multiplying that figure with 78 km (i.e. the difference between the mileage Winnipeg-Regina-Saskatoon and Winnipeg-Melville-Saskatoon: 837-759=78) and dividing it by 4466 km (i.e. the total distance on the route of today's Canadian), we can estimate that the incremental costs of rerouting the Canadian between Winnipeg and Saskatoon via Regina (rather than Melville) would be $1.2 million.
To estimate the impact on ridership, let's take the annual ridership of Saskatoon (
4,289 passengers boarding or deboarding in a city of 295,095 according to the 2016 Census) and adjust it proportionally to reflect the size of Regina and Brandon (i.e. the only two CAs or CMAs which would gain service by this re-route):
Regina
Population: 236,481 (i.e. 80.1% of Saskatoon)
Estimated ridership potential: 3,437 (i.e. 80.1% of Saskatoon's passenger count in 2018)
Brandon
Population: 58,003 (i.e. 19.7% of Saskatoon)
Estimated ridership potential: 843 (i.e. 19.7% of Saskatoon's passenger count in 2018)
This suggests that re-routing the Canadian through Regina and Brandon might gain 4,280 passengers, which means that every single of these 4,280 passengers would need to pay $272 to offset the incremental costs of the re-route. However, the re-routing would deprive Rivers, Melville and Watrous of their passenger rail service, which accounted for 722, 1438 and 354 passengers, respectively. If we deduct these 2,514 passengers, we only have a net increase of 1,766 passengers and the break-even ticket price would now rise to $660. Given that even a full-fare "Economy Plus" ticket between Saskatoon and Winnipeg (i.e. a longer distance than Regina-Winnipeg or Brandon-Winnipeg) costs only $188 while the cheapest fare is $106 (I checked for May 2021), I find it highly unlikely that such a high average fare could be achieved in such small cities which are mostly irrelevant to the kind of passengers which ride the Canadian in Sleeper Class.
It’s this kind of back-of-the-envelop analysis which makes me doubt that re-routing the Canadian onto a longer route (and with partly using branch lines) would be possible without increasing its need for operational subsidies...