Demographics matter a lot in some circumstances. I don't know that they matter as much in such a low-turnout race. Compare Democratic primary turnout in 2016 to 2008 (or 2012, for that matter). It's not very high. There's a lot of selection going on. It doesn't matter if Hillary leads in a demographic by 10% if that 10% margin fail to turn out. Getting those people to turn out for Hillary is even more difficult when you look at her trustworthiness scores. This is why I'm irrationally hopeful for a Draft Warren/Biden candidacy.
You do indeed seem rather prone to irrationality ... I understand fully why Sanders is attractive to white -- and let's be real here -- relatively privileged people. But I strongly doubt he'll find the same sort of success with non-white Democratic voters. You seem to be completely discounting just how demographically favorable Iowa and New Hampshire were to Sanders. If Sanders were to prove that he can win in massively more diverse primaries and caucuses, then I'd take him more seriously as having a path to the nomination. But so far he hasn't done shit to prove that. You are giving him credit that he has not earned.
The only way I compared this election to 2008 is in the PUMA voters (that never really existed). Other than that, you're making points against an argument that I'm not making. I created a hypothetical that I noted was unlikely, which is your fifth point there. I'm pretty sure we're in 100% agreement on this.
Hillary has been performing better among people earning more money. Why would Sanders's message appeal to people who are going to be paying a significantly higher percentage of taxes under a Sanders administration?
What I'm saying is that yes, NH and IA were pro-Sanders in a lot of ways. But the primary electorate so far is shaping up to be more beneficial to Sanders than Hillary. Obama was able to mobilize African-Americans to vote in large numbers in the primaries. Hillary isn't going to have that same ability. If she leads Sanders among African-Americans 70-30, but turnout for them stays low in the primary (not the general - African-Americans will most likely be the most mobilized group in 2016 for structural and sociological reasons), that advantage will be less worthwhile than if Sanders boosts privileged white kids in those same states. Low-income white voters don't turn out in lots of elections as-is, and if turnout stays depressed, Hillary's margins of victory will continue to underperform polling. Sanders was supposed to lose Iowa by 3% and win NH by 17%, according to 538. He lost Iowa by 0.25% and won NH by 21%. We only have two states, so he could just be overperforming "at random," but it's plausible to me that Sanders voters are mobilizing whereas Hillary voters are not. Demographics don't mean jack if turnout is selective.
And yet, even by your own account, she wins. You are reading so much into two states. Two states that are notoriously hard to poll. Two states where campaigns have months (even years!) to mobilize voters. Of course Sanders mobilized his voters in those states. First, those voters existed in disproportionately high numbers. Second, they were extremely well organized. Like I said, he still hasn't proven anything beyond the ability to wallop Clinton in an extremely atypical, demographically favorable state next to his own home state, and an ability to nearly draw Clinton in another extremely atypical, demographically favorable state. Missouri is one of the few states I've never visited. But as far as I'm concerned, when it comes to this primary season, I'm in the Show Me State. And Bernie hasn't.
That's what we're gonna need Nevada and South Carolina to show us. We haven't had a new poll for either race in two weeks. This lack of certainty is what causes many an irrelevant debate.
You probably won't get a poll for Nevada. And certainly not an accurate one. Like I said, those caucuses are shady as hell. I'll just be seeing what Jon Ralston says ... https://twitter.com/RalstonReports