I did a recalculation including the SES numbers. The only shift is the switch of two NDP seats to the Liberals to give them 78 and the NDP 37. SES has generally found slightly better results for the Liberals and slightly weaker numbers for the NDP in this campaign.
There are between 55 and 60 seats (depending on what poll numbers one uses) where the margin is 5% or less. That is a lot of uncertainty.
I have noticed in other elections a small tendency for the NDP to fall a bit short (I am not including 2004 here) of their poll numbers on voting day. My guess is that it is pollsters oversampling younger voters who then don't vote. It is a small effect in any event.
The NDP is definitely going to do better this time, perhaps much better.
In Ontario, based on the forecast that includes the SES numbers, I have them winning all their current seats plus Algoma-Manitoulan, Hamilton East-Stoney Creek, Hamilton Mountain, Kenora, London Fanshawe, Nickel Belt, Parkdale-High Park, the two Thunder Bay seats, Trinity Spadina and Welland. There are three other seats where the NDP is within a point or two of winning seats from the Liberals - Beaches East York, Davenport and Sudbury. They are also close to taking Oshawa from the Conservatives. With one of the stronger new NDP candidates running in Beaches, Marilyn Churley, I would guess that is likely to tilt the NDP's way. Many of these races are close and a closing tick to the Liberals could still tilt some of the ones I have going NDP back to the Liberals.
In BC I have the NDP gains as New Westminster-Coquitlam, Southern Interior, Surrey North, Vancouver Kingsway, Vancouver Island North, and Victoria. I think they could also take Esquimalt-Juan de Fuca but my numbers show the Liberals hanging on there. They are also close, as are the Liberals, in Newton-North Delta. I see about nine very close races in B.C.
When one applies a mathematical projection model, "close" means we are essentially describing races too close to call. Same goes for many of the races noted above in Ontario.