Are you saying that the #analysis will be that it's gone for Democrats in 2020? Because the idea that Republicans can't win PA is definitely dead by now.
No, for Republicans. I've already seen like two people here (most notably pbrower2a) saying that 2016 was a fluke and that the state will go back to being "likely D" in 2020 or whatever based on that one poll that showed him with poor approval ratings in the state. PA might vote Democratic (and I agree that it's the most or second-most likely state to flip to the Ds), but I definitely think it is winnable for Republicans (the same way FL was winnable for Obama in 2008, if you want).
But generally the "fools gold" analysis isn't very useful in most cases. Many states are always fool's gold until they aren't. Democrats could win AZ, just like Republicans could win MN.
Edit: Another thing that could be rejected is the idea that <candidate X is too liberal to be elected nationwide!!>.
I agree. The term "fools' gold" should be retired from political discussion (sure, there are enough solid Democrats in Colorado at this point to make the state vote consistently Democratic, but they still could decide not to show up), as should the idea that voters run all the candidates through Political Matrix and decide based on that who they're allowed to consider.