I can't speak for others, but for me, a big factor is the amount of transit that is provided. Surface LRT, even in it's own ROW (albeit with at-grade cross-streets) is a fraction of the cost of subways. People can argue ad-infinitum about projected demands and whether they are for subway capacity or LRT capacity, but if you buy in to the projections that are only to the latter, then it makes far more sense to build four (or more) times LRT kms than subway kms.
Not neccesarily. The $4.6 billion being used now to do a mere 20 kilometres of Eglinton Crosstown LRT at roughly $230 million/km is enough to build 15 kilometres of Eglinton heavy-rail metro (@ $310 miilion/km). These costings can flactuate even lower the further west one expands that subway. So a subway stretching from Highway 27 and Dixon to Yonge and Eglinton
or one from Kennedy Station to Eglinton West
or one from Mount Dennis to Warden could be built for the same amount and carry tens of thousands more people during AM/PM rush than the light-rail ever will be capable of doing. Can you honestly say that LRT is the better bang for buck in this situation? Or consider FWLRT's $1.2 billion for 11 kilometres with over a billion dollars more expense if they ever extend it across to Don Mills. BRT costing as low as $15-25 million per kilometre could give us a line stretching all the way to the Zoo.
What Transit City pundits also fail to take into account is the fact that the average North American light rail line has run 35.8% over budget:
Cost Overruns and Demand Shortfalls in Urban Rail and Other Infrastructure-
http://www.sbs.ox.ac.uk/centres/bt/Documents/URBANRAIL6.1PRINT.pdf
The design life of a subway is +100 years. You don't build a subway for the needs of today, you build it for the needs of tomorrow. The expense to build a subway in an area that is already "at capacity" as you suggest would be vastly more than what it is now. Plus you would have the inconvenience of shutting down a LRT line that's at full capacity. Basically the Sheppard Subway won't justify it's existence for 10 or 20 years, but prove invaluable for 80-90 years after that. Cities need long term plans, not short term stop-gap measures. That is, the opportunity cost of not installing subways before the area around the Eglinton corridor intensifies is insurmountable.
That being said, I'd give preference to the DRL before any TC lines that will be dumping more people on an already at-capacity subway line. However I don't see how a new mayor coming in and outright canceling TC will suddenly secure the necessary funding to build the DRL, especially since it isn't high on the priorities of any of the anti-TC candidates.
And unfortunately it's thanks to people like Jane Jacobs who fought to keep streetcars running through the Downtown core that the construction of the DRL has been put off for over 40 years now. Without streetcars that subway would have been expedited a long time ago. No, I'm not saying that existing lines need to go before I get misquoted (again); but certainly the 501 and 504 corridors are in serious need of alleviation. To my understanding the Metrolinx board is already highly in favor of
subways along Eglinton and Queen/King and have placed these lines on their earlier press releases. That's half the journey right there. They have also expressed interest in running the Eglinton Line up to Scarborough Ctr, to give it a major destination to terminate at. With the right Mayor and City Council in place backing the pro-subway agenda, I could also see Bloor-Danforth getting extended as part of an interline during the same ROW construction.
A bus-based rapid transit system for other parts of the city is also doable in a lot less time than it'll take to build TC. Busways are convertible to light-rail corridors in the future as demand grows. At 2800 PPH during peak hour by 2031, neither Finch West nor Sheppard East can quality for light-rail based on the TTC own stats regarding carrying capacity. Bus lanes can take up a maximum of 20 metres of roadway but in many cases that much width may not even be warranted (mainly just at the stations for bypassing). That it only took 18 months to complete the York U busway while the end-date of construction for the St Clair right-of-way which started over 5 years ago is yet to be determined is telling of which mode produces faster results.
So in summary, sometimes change is good - change in gov't, change in priorities. Ottawa holded off building its LRT network for a few extra years and now they're coming out with a far superior plan which involves extensive grade separation to keep pedestrians safe and warm and commute times actually fast.