I would not put too much faith in those polls. First, the election is too far off for the polls to be meaningful. Second, polls one month before the election in 2016 showed Clinton with a larger lead - and we all know how that ended. Just a few examples from October, 2106:
NBCNews: Clinton 50% Trump 40%
Bloomberg: Clinton 49% Trump 37%
Quinnipiac: Clinton 52% Trump 44%
IPSOS/Reuters: Clinton 44% Trump 37%
Fox News: Clinton 49% Trump 41%
AP/GfK: Clinton 54% Trump 41%
USAToday: Clinton 49% Trump 39%
On November 8, 2016, CNN gave Clinton a 91% of winning the election. The same day, Nate Silvers at 538 predicted that Clinton would win Fla, NC, PA, MI, WI, and NH. She lost all but NH, which she won by 3,000 votes. IBD/TIPP was the only outfit that predicted a Trump victory. IBD/TIPP also tracks economic optimism; which was high in 2004, very low in 2008, turned positive just in time for the 2012 election, and was slightly negative in 2012. It hit a record high recently. In the words of Bill Clinton, "it's the economy, stupid."