The quoted post is from the "Downtown Core Line - Possible Alignments?" thread, but really belongs here.
... there is NO reason TO extend Sheppard anywhere because the ridership is far too low... FAR too low. It's a money pit. Extending it will only make it a bigger money pit. Really stupid use of tax dollars to provide capacity that isn't needed.
That cost argument would apply to the original TC cost estimate for the Sheppard E line (560 m). Invested in the subway, that amount would only get the subway from Don Mills to VP.
However, the cost of Sheppard East related projects currently on the table is quite a bit higher: revised 800 m for the Sheppard E line, plus at least 300 m for the Finch / Don Mills LRT bypass (and perhaps more, if it has to be partly tunneled). For that amount of money, the subway could probably get from Don Mills to Kennedy.
The LRT is a far better solution to this. It's worth pointing out that even the LRT is on the low end of demand, only 3,000.
The same model that predicts 3,000 for Sheppard E LRT, predicts 5,000 for the subway extension. This is still low by subway standards, but there are two other factors in favor of the subway extension:
1) There is no alternative E-W rapid transit corridor across the north of 416, while the city is wider there than in the south due to the lake's geometry. A subway from Downsview to Kennedy can serve as such corridor. In particular, Finch E and Ellesmere bus routes (or future LRT lines) could be operating off the subway terminus at Sheppard / Kennedy. In contrast, the proposed Sheppard subway / LRT combo won't be useful for riders living / working at Finch E or Ellesmere.
2) It is difficult to schedule multiple branches of the planned Sheppard LRT (for example, Meadowvale / Zoo, Scarborough Centre, and Warden North). The total demand is within LRT capacity, but the problem is that Roads department does not want LR vehicles to run more often than once in 5 min, to avoid blocking the general traffic by LRT signal priority. So, to enable branches, we would have to either settle for very infrequent service on the branches, or run a very frequent service (2 or 3 min) in the common section but forfeit the signal priority. Obviously, the subway extension would not suffer from this problem.
Therefore, I believe that there is a case for extending Sheppard subway at least to Kennedy in the east, despite the relatively low ridership projections.
Eglinton will about double that, for example, and Eglinton's ridership is also still too low for a full-blown subway. LRT is the best solution for these corridors and the best use of limited funds.
For Eglinton, I agree that LRT is the best solution. The ridership projections for the Eglinton and Sheppard corridors are similar (perhaps Eglinton is even a little higher), but there are other routes for rapid E-W trips across the middle of 416: Bloor subway, and (hopefully) the future GO service on the CP's Midtown line.
... both Eglinton and Finch will provide the exact same function [connecting the Yonge and Spadina subways], both of which will be (mostly) underground (Finch is probably too narrow east of Bathurst). Why do we need a subway between two LRT lines that can handle the traffic between the two perfectly well?
The relevant section of Eglinton will be underground, but quite busy. Finch between Yonge and Keele will be mostly at surface (only a short stretch of Finch W, perhaps 300 m just west of Yonge, looks too narrow for surface LRT). So, a subway link between Downsview and Yonge / Sheppard won't be redundant.
Other advantages of extending Sheppard subway to Downsview would be a much better connection from York U to north-east, and from north-west to north-east (for example, from Keele / Wilson take a bus to Wilson stn, subway to Downsview, another subway to Warden North, and a bus to Finch / Warden).
Yet another bonus of the subway link would be non-revenue movement of trains.
I don't think that extending Sheppard subway to Downsview is urgent. It does not have to happen within the 25-year plan, but should be done at some point in future.