Jamelle Bouie argues today that Republicans may well win the presidency without actually changing much or offering new ideas.
I think that's broadly correct: the out-party just doesn't matter all that much. I'd put two caveats on it. One is that the GOP crazy has caused problems in their efforts to find an appealing general election candidate. It doesn't matter a lot, but it matters some, and they'll probably have to find some way around it or suffer an electoral penalty (which, to be sure, would be on the margins and probably wouldn't make much of a difference). The other is that actively annoying large swing constituencies is also something that can hurt them and may have hurt them this year. You would think it would be fairly easy to cut that out (avoid rhetoric that offends Latinos! Don't call women sluts if they advocate for policies you don't like!), but a good deal of this is structural, and thus harder to end than one might think. That is, if very visible Republicans have (financial) incentives to do that sort of thing, it may be difficult for electorally-oriented Republicans to do much about it.
Again, however, we're talking somewhere below five percentage points total from all that, presumably, and probably a lot less. It's unlikely to swing the election. And both of these are the sorts of things that may well be less evident after eight years out of the White House -- not that the causes will be cured, but that everyone will be willing to play nice for the duration of the campaign.
What I would say about it, however, is that the dysfunction in the current GOP makes successful governing if they do win extremely difficult. I think we've seen that for some time, and I think it was part of why George W. Bush was such a poor president; there really are major governing penalties for finding it hard to accept reality.
Put it this way: it seems that Republicans outside of the recent campaign certainly deluded themselves about the polling, and Republicans inside the campaign may well have been guilty of it too. I've argued that it didn't really hurt them, however; mostly it just meant that the eventual results took them by surprise. However, if you try to govern that way -- say, if you actually believe phony revenue estimates, or actually believe that people in some country are eager for you to invade them when in fact they are not -- then massive policy disasters are likely.
As I've said many times, this doesn't really fit on a conservative/liberal or moderate/extreme framework all that well. What's important is being able to get key clues about whether policies will work or not, clues that for a normal, healthy party are produced by the normal political system. To the extent that the current GOP is so dysfunctional that its politicians are likely to ignore those clues, or will structure their administrations so that they won't be able to hear them, the chances for catastrophic policy failure go way, way, up.