Hardly anybody mentions the undecided voters, which gives a false sense of what the polls are really saying. It's more revealing to look at the numbers for all voters:
Trumpeting only the figures for decided voters, as most of the media do, is equivalent to distributing the undecided voters over the parties in the same percentages as the decided voters, which ain't necessarily so. There are enough undecided voters to give a thumping big majority to the Conservatives if they all vote that way, at least a solid majority to the Liberals if they all vote that way, and even a shot at official opposition status to the NDP if they all go left. Highly improbable outcomes, I know, but if the undecided voters don't split approximately the same way the decided voters do, and there's no reason to assume they will, all those pollsters can end up looking pretty silly. And of course we all know the BQ numbers apply only in Quebec. As Jersay observed, it's still up for grabs.
It's still too close to call because of that big undecided vote. Also, if this is the standard "plus or minus 3 percentage points 19 times out of 20" sample, that means the pollsters are 95% sure the actual Conservative vote is somewhere between 28 and 34%, the Liberal vote is between 23.5 and 29.5%, and so forth. And the undecided vote might be as high as 18.9%. Plus there's a 5% chance the numbers are what a former high school teacher of mine would have called concentrated essence of hogwash, so don't bet the farm on anything yet.