That's the same poll that has the Tories on 41% of the vote, which is 5% more than they polled in the general election, so there's not much to be happy about that there either. (Labour increased by only 3% btw)
Besides that misses a key point, which is that the Labour Party almost always gets more popular between elections than it does at general elections themselves. This
excellent blog post looks at it in detail, but in a nutshell, only once in history has the Labour Party done better in an election than they were doing at this point, six months into the term (that was Blair in 1997 when everything went right for Labour). On average they lose 6.1 points at the polls compared to this point. Thats when in opposition. If you include their times in Government, they lose 8.4 points.
The Government also gets less popular mid-term, but more popular towards a GE. On average among all the parties, the Government gains an average of 3 points as the election grows near. Its called the swing back tendency.
Its a bit like judging the final result based on the score midway though. If its 0-0 with 10 minutes gone, you can't draw any conclusions. But if you're 3-0 down with 20 minutes gone, you can assume the game is lost.
The fact is that, judging from the polls (and, frankly, from this observer's judgement) Labour are headed to their worst ever defeat in 2020 and sub-30% is very likely.