I've seen people mention a negative trend for the PPC, but that's misleading. They literally took the order of polls on Wikipedia as chronological, even though some are on the same day from different pollsters. If you separate the polls based on who actually performed them, removing the bias between pollsters, you get a positive trend.
Nanos Research latest polling gives the PPC a 5.3% share of the vote. This compares to their last poll from the day before, that gave them a 4.3% share. The She-covery Party, by comparison, dropped 1%, while the CPC retained their share of the vote from the prior polling.
Mainstreet Research, with a poll from the 8th September, gave the PPC 7.9% of the vote share, which is a +0.3% difference from the last poll. Unfortunately, the She-covery Party gained 2.3% in the polls as well, moving from 30% to 32.3%. The CPC lost 0.6%.
Finally, the EKOS polls gave the PPC a 9.5% share of the vote. This is up from 7.9% in the last poll taken on September 6th. The She-covery Party gained 1% in that period, while the CPC gained 0.6%.
(She-covery Party is a joke at Trudeau's expense, it is of course the LPC)
Going back to a month ago, these same pollsters gave these figures. The newer ones are in brackets for comparison
Nanos - Aug 20th :
CPC - 32.3% (32.6%) +0.3% since August
LPC - 34.2% (30.6%) -3.6% since August
PPC - 2.1% (5.3%) +3.2% since August
Mainstreet - Aug 15th :
CPC - 30.1% (32.9%) +2.8% since August
LPC - 33.1% (32.3%) -0.8% since August
PPC - 5.6% (7.9%) +2.3% since August
EKOS - Aug 15th :
CPC - 30.3% (34.3%) +4% since August
LPC - 34.8% (31.8%) -3% since August
PPC - 5.1% (9.5%) +4.4% since August
TL;DR : A PPC victory is possible, but is there enough time for the momentum to shift entirely? The polls could be being manipulated to make it look like the CPC is the only way to stop Trudeau winning. The growth of the PPC doesn't appear to come at the expense of the CPC, as they both grow in projected vote share around the same rate. Of course, it could be LPC voters switching to CPC and then loyal CPC voters switching to PPC. This trend would be interesting to have information on.
I believe the CPC will win though, because of people who don't do much research and vote for the same party because they think it still represents them, and the main opposition being so useless it makes their main voting block turn against them. (See UK Conservatives for an example of this.)
How did I do, Canadians? Was this valuable?
I could do deeper analysis, this just took polls from two consecutive days in September and then one from the beginning of campaigning to try and work out the short-term and long-term trends.