I think the incentives toward accuracy are pretty decent. It's easy to compare polls to one another and also to compare them against the actual result. And pollsters make a lot of claims about their accuracy when soliciting new business. Plus there's professional pride, etc.
You do encounter issues that also occur elsewhere in political media. For many outlets, most of the audience will consist of political partisans. Even for these outlets, though, polling is sometimes an avenue they use to bolster their claims of unbiasedness and objectivity.
For example, whatever else you think about Fox News, their polls are both very good and quite unbiased (if anything they've had a tiny D bias in recent elections). projects.fivethirtyeight.com/pollster-ratin…
Maybe there are some "post-truth" outlets that don't even care about the pretense of objectivity. Even then, putting out inaccurate polling is pretty risky as it could serve to demonstrate to otherwise-credulous readers how disconnected you are from reality.
There are exceptions, the most obvious being Rasmussen Reports. Their brand equity is tied up in The Poll That Always Has Good News For Republicans. And they thrive off the ~30% of elections where polls have a D bias and they look smart (& hope people forget about the other 70%).
