Ohio wasn't some massive swing towards Trump; Clinton hemorrhaged former Obama supporters, due to a variety of reasons. Let's take a look at turnout in the state from the 2012 and 2016 Presidential Elections. Overall, 54,387 fewer people voted in 2016 than 2012.
5,590,934 - 5,536,547 = -54,387
Now, look at the change in raw numbers from Obama to Clinton and from Romney to Trump. There were 433,540 fewer people who voted for the Democrat in 2016 than 2012, but only 179,569 more that voted for the Republican.
2,827,709 - 2,394,169 = -433,540
2,661,437 - 2,841,006 = +179,569
Basically, Trump experienced a decent increase in votes over Romney, but Clinton experienced a dramatic decline over Obama - especially from his 2008 peak, and even from Kerry's results. There's nothing to indicate that (a) there were a significant number of Obama-Trump voters and (b) the next Democrat cannot recreate the results that Obama received simply by turning out the vote.
Like I was saying to Virginia in a different thread, Trump staying steady while Clinton dropping heavily from 2012 doesn't mean it was just a direct transfer from Obama → 3rd Party. And in this case, Trump actually got 3.7% more than Romney.
And you're points about turnout ignore the fact that Ohio is a state with very low growth, and a large chunk of it is actually losing population. The more frightening point is that, despite that, Trump still got more votes than Obama 2012. If you look at the county results, he clearly held Romney's voters together
and added a significant number of Obama voters across the state but particularly in NE Ohio.
This is a lot more daunting a scenario than you think, IMO.
