No, he's not using the numbers fairly. There is no way to fairly compare final project costs, after delays and overruns, with a rough estimate (and Transit City's estimated cost has already been revised upwards by $2 billion). The first example he uses to try to prove that subways are "too expensive" is Spadina, and he's using wrong numbers. TTC estimates for subway projects are all ridiculously high; they're all 100% tunnelled and as gold-plated as possible and do not include the slightest attempts to save costs...and why would they make such an attempt when they can tout an "affordable alternative" like LRT?
Why do these "subways must be X times as expensive as LRT" figures keep spreading around the internet even though they're never applicable to real projects? They're just another chapter from the "LRT is our messiah" Bible. It makes absolutely no sense that a subway in a shallow trench would be cheaper than LRT in a tunnel, but some guy on the internet said it is, so he must be right. The difference between fully tunnelled LRT and fully tunnelled subway is marginal...marginally wider tunnels and marginally longer platforms and a few other marginal costs. But we won't know how small this margin is because they're not studying the option. If they actually studied it and concluded that the ridership wasn't there even in best case scenarios and that it wasn't worth spending $X, fine.
1) Calvin Henry-Cottam's cost estimates for LRT are based mostly on the LRT projects already completed in other cities, rather than on the unproven Toronto's Transit City prospects. He made an honest attempt to produce best estimates, based on the available data.
2) His subway cost estimates are not actually based on the Spadina extension. On the contrary, he mentioned that Spadina is abnormally expensive. All his estimates (200 - 250 million per km for underground line, 150 - 200 for surface) come lower than the TTC's Spadina plan (291 million per km).
If he just wanted to bash subways rather than assemble accurate data, he could use that 291 million per km figure to raise his upper margin for underground subway, and give a range of 200 to 300 million. But he did not.
3) Obviously, LRT is not a magic universal solution for any transit route. It has its strengths (cost and construction speed) and weaknesses (capacity and travel speed) compared to an HRT line.
The biggest cost saving comes in cases where LRT can be on-road (median lanes or side) while HRT must be elevated, trenched, or underground. For such routes, both the estimates and the simple common sense suggest that the saving will be several-fold.
For the sections where both LRT and HRT are underground, or both are elevated, obviously the saving factor will be much smaller. But I think that it will be more than marginal. Not sure about the impact of tunnel width. But a platform for 6 cars is not "marginally" shorter than the one for 3 cars; it is two times shorter.
4) I have to agree with you that TTC does not strive to build its lines in the most efficient manner, and that applies to both HRT and LRT. Unfortunately, I have no clue how to fix that. Vote down all Chancellors who sit on TTC board during the next elections? But who can voucher that the new board will be more effective? Plus, the senior engineering staff will remain unchanged anyway.