I'd call all of that as pretty weak evidence when there's no trend in party affiliations in the last 20 years. Looking at a change over a couple months (for approval ratings) and some easily explained away exit poll differences doesn't really convince me. Here's some pew research data if you're curious.
If they do this analysis again in 4 months and there's some kind of change in the party affiliation data then I'll agree with you, until then you are just seeing what you want to see based off next to nothing.
edit: that said, I appreciate you giving me some kind of evidence and not just making the claim over and over again, even if I find the evidence weak.Last edited by Torillian - on 10 November 2018