It should definitely make us all nervous, but lemme try to help out just a little bit:
The principle behind 538 is that no one poll all that useful as a prediction tool, that it's better to aggregate all the polling information to come up with a few of where things are going. They used to be the only ones doing that, but now there's others. And I would argue that the same principle should apply to aggregators who analyze all the data and trends- you get a better picture if you look at all of them. Of them, 538 is the lowest on Clinton's chances at the moment. You can see them all
here about halfway down the page in "win probability." The others all have her between 63% and 81%. The one with 81% is the Princeton Election Consortium, which has a very good reputation.
And as you'd expect,
the betting markets put the numbers somewhere in the middle of all this too, making Clinton about a 2-1 favorite (66%). Although obviously they can also account for expected debate performances, or the increasingly slim possibility of America finally pulling its head out of its collective ###.