West Virginia's been long gone for Democrats since 2004. Even if Hillary doesn't lose it by 32 percent, I doubt she'll come within 20 points of the Republican. I've always thought that these "Candidate X is a good/bad fit for State Y" arguments are overly simplistic. Hillary won't significantly overperform Obama in West Virginia, Arkansas, or Kentucky (unless she wins in a landslide), and she won't significantly underperform Obama in Colorado (unless she loses badly.)
Depends on your definition of significant. The state going from a 5 point D win to a 2 or 3 point R win, which I think is a reasonable CO result even if Hillary wins nationwide, seems pretty significant to me.