I already mentioned something similar could be done east of Leaside (edit - because, of course, why would design choices that consciously streamline a project be done on only part of a line?), and the line could be not tunnelled beneath Eglinton Flats, too. Throw in some elevated parts and cut'n'cover segments (keeping the line as close to street level as possible at all times) and we're looking at more like 15+km, and if the line went from Markington to Pearson it'd be more like 20km.
If Eglinton is built as a subway, tunneling will be needed, at the very least, from Weston Rd. to Leslie (11 km), from Don Mills to Pharmacy (3 km), and from Warden to Kennedy (2 km), total 16 km. Estimating 16 km at 300 M / km plus remaining 14 km at 200 M / km, the cost goes to 7.6 B, versus the 3.5 B estimate for the LRT solution.
Surely the cost depends on the engineering complexity rather than just on the length, but the Eglinton subway would have its fair share of complexities. It would have to cross Humber River, West Don, and East Don (LRT would have to cross them, too, but it might fit on the existing bridges). A connection to the existing system at Yonge or at Eglinton West would be difficult to build because the areas are built up. (Connection at Kennedy would be easy, but starting the Eglinton line from there is quite odd).
A successful push for the Eglinton subway will have one of the following outcomes:
1) Eglinton subway gets build fully, but costs at least 3 or 4 B more than the light rail solution, and that expense derails a number of other transit projects in GTA. It doesn't mean just derailing Jane or Morningside LRT (that are not likely to get built anyway), but really important projects like Downtown Core line.
Or
2) Only the central section (likely Jane to Don Mills) gets built within 25 years, and bus service is retained at the extremities. The line could be extended later on, but in the interim, we'll have a pretty poorly connected system. A stubway on Eglinton, another stubway on Sheppard (hopefully lengthened but likely not reaching the full project length). A Downtown Core stubway - that one would see a very decent usage even at the stubway phase, but anyway.
Regarding the LRT solution, I don't see a big problem if it runs in a 2 B dollar tunnel in the centre, but stops at red lights at the extremities. This is how many other LRT systems in the world operate.
However, I'd extend the fully grade-separate LRT section a bit, and make it Jane to Don Mills rather than just Weston to Laird. If that section is built properly (can handle long trains and allows tight headways), its capacity limit could be in the 15,000 - 20,000 pphpd range. This is more than enough to support local passengers (even if densities grow) as well as transfer passengers. Long 4-car or 5-car trains would operate between Jane to Don Mills only, whereas 2-car trains would run end to end.
Sections west of Jane and east of Don Mills would be serviced by 2-car trains on 4 or 5 min headways, which is still a lot more capacity than what regular bus routes typically provide.