you realize that some of those seats - Fort McMurray-Athabasca, for example - have literally *never* elected a non-conservative, right?

there's a wide swath of alberta, saskatchewan and rural ontario that will never vote for anything except the most right-wing major party. the conservatives take these seats for granted. there's something like 100 of them. but, if that truth were an insurmountable advantage, the conservatives would not have spent most of the last century in opposition.

there's a lot of truth in the idea you're working with, which is that regional support does not translate well into national numbers. but, if you look at the regional breakdowns, what that suggests is that it's the liberals that are in extreme trouble because they don't have a regional base.

that 34% should be plenty to allow them to win, if it means strong showings in quebec & bc and a reasonable showing in urban ontario. but if the liberals end up below the ndp across the country, they could pull in 30% and get 30 seats; they could finish in second place in 300 ridings. and, while that seems extreme, it's my current projection given the existing numbers.

as you say, the conservatives will get their safe seats. but that's not a majority. and the ndp will take just about everything else.

..so long as they can continue to hover around 35%.

thetyee.ca/Opinion/2015/08/15/Note-New-Dem-Strategists-Dig-Deeper-Polls/