As commuters move further away, the lines will get extended. They did it to Kitchener and to Barrie. So how would they extend the line to Orillia. It will be the next place to extend North.
And what tracks would they use, exactly, to service Orillia as an extension of the existing Barrie line? I'm not aware of any tracks existing in or immediately near Orillia at all, and certainly not as an extension of the tracks at Allandale Waterfront.
There were existing rails to Barrie South and then Allandale Waterfront, running service was just a matter of fixing a few things up, building stations, and having the trains continue after Bradford.
For Orillia, all-new tracks would need to be constructed, which would be unprecedented in GO's recent history, wherein they have only extended service along existing rails or they've created/are creating (e.g. Oshawa split service) a small connector rail to move trains onto a nearby existing line from their current route.
Also, I'm not sure where you'd want to put those tracks. Allandale is roughly in downtown Barrie, you can't demolish their downtown to install a railway. You might run trains west-southwest along the existing tracks then curve around just outside of Barrie's city limits to head northeast to Orillia, but again this would require unprecedented new track construction/right of way acquisition, and result in a long circuitous route to get to a small town.
Same thing for Kitchener. There were existing rails and even existing VIA stations at Guelph and Kitchener, fairly little infrastructure had to be built for service to be extended, relative to what already existed there.
The other GO lines can all extend, subject to negotiations with CN/CP--LSW is expanding towards full(er) service through to Niagara Falls (where it must terminate due to the US border) and could gain another branch towards Brantford/London if ever justified, Milton to Cambridge, Kitchener to Stratford, Richmond Hill to the middle of nowhere, Stouffville to Uxbridge, and LSE to courtice/bowmanville/clarington then if justified to Port Hope, Coburg, and through to even Kingston. However, such expansions are beyond GO's current mandate, and in the LSW/Kitchener/LSE branches are much better served by existing VIA services.
At some point, it will become impractical to continue expanding GO lines to the point that it's a 3, 4, 5 hour trip from Toronto.
Look at Bloomington GO station as an example of Metrolinx's big-picture, long-term strategy: build a station far out on a GO line, adjacent to a highway/major commuting route, with tons of parking, and have it be very attractive, all in order to get people coming from far away on that highway to get off at that exit, and switch to the train instead of continuing downtown in the car. So far from Toronto, there is ample excess highway capacity and it largely makes more sense to expand highways than rail service due to the spread-out small communities. For instance, Bloomington can capture commuters from the southeastern side of Lake Simcoe in Keswick, Beaverton, etc., who can drive down the 404 and switch to GO at Bloomington, rather than building new rails up there for a few riders; there's excess capacity on the northern 404 so having them in cars up there is fine, but this gets them out of their cars before the congested Richmond Hill/Markham/Toronto stretch.
A lot of anti-GO-parking-garage people are making the same fallacy saying we should improve local transit service but maintain a general status quo of traditional buses serving traditional routes; it simply won't provide much benefit for the cost, and to provide substantial benefit we'd need to run so many additional bus routes it would be prohibitively expensive. In that case, we need to go down some sort of micro-transit uber-like service, much as in the rural areas we need partial-drive partial-train commutes rather than full-drive or full-train.