1.) We are never going to have both a Queen and DRL subway. I can't really stress enough how unlikely that is to happen. It is fairly rare to build parallel subways less than a km apart from each other in subway mad jurisdictions like Madrid. In Toronto, it simply wont happen. Ever.
2.) There are routing difficulties involved with building both a Queen & DRL subway. The DRL "U" makes sense. Once that is built, a Queen subway would likely turn into a strait line following the existing streetcar route. That would be overly expensive and unnecessary (Nevile Park does not need subway). The only part of Queen that justifies a subway is the central segment, roughly between Pape and High Park.
3,) Why not design a route to be compatible with branch lines? The reality is no matter how you cut it, no one line is going to achieve adequate coverage of the City. The best way to deal with this is to build the odd spur line to meet a specific objective, not holding out for a separate parallel subway to be built.
4.) You can't ignore other projects in the area. Queen's Quay will likely have a well functioning LRT no more than 300m from the rail corridor. Depending on execution, this should serve much of the waterfront development fairly well. Also, GO Train improvements should eventually bring headways of less than five minutes in between the Don River and Exhibition. With improvements to the Ex station (to serve Liberty Village) and a possible Cherry Station (to serve the WDL, EB ect...), a good deal of commuter traffic can be accommodated. It seems redundant to add another rapid transit line to the mix.
5.) Queen/King can't easily, and most likely wont, be converted to anything approaching LRT. Space is just too tight and local merchants seem totally against it. In the longterm, something will have to improve capacity on the downtown part of this corridor. The only realistic solution to that, in my mind, is a subway of sorts.
6.) Zoning can be changed. Subways are multi-billion dollar investments that last for centuries. It makes no sense to base them on completely flexible zoning policies.
7.) Adelaide or Richmond would be better than King or Queen. They have no revenue public transit service, relatively light street retail presence and therefore easier to build a subway along. Ideally we should cut/cover this for as much of its length as possible. Regardless, it would probably be cheaper to just temporarily shut down Adelaide/Richmond than trying to build a subway on Queen/King while keeping things business as usual.
8.) The North Eastern arm of the DRL (i.e. Pape-Eglinton) is more important than the western arm (i.e. Spadina-Dundas West).
9.) How much of the rail corridor is feasible to build in? It is cramped east of the Don River, cramped west of Union. That implies only like 2km of feasible ROW. That simply wouldn't have a noticeable affect on costs. The only leg I can see that really makes sense is in the West, up from Parkdale to Dundas West.
10.) Keeping those in mind,
something like this would probably be the best alignment possible to minimize costs and maximize ridership. Really, if we ever did get both ends of the DRL "U" up to Eglinton, it would seem logical to then just close it off along Eglinton and make it a circle.