Bernie could take 100% in the remaining caucus states, and if he doesn't do well in the big primary states, he'll still lose by 500.
Remaining Caucus States: Wyoming, Guam, US Virgin Islands, Puerto Rico, North Dakota... Total 132 delegates if he takes 100%. He won't, but assume Alaska-like figures if you wish... 80% = 105 Sanders : 27 Hillary.
That's about the lunch-time allocation in New York. His requirement after today may have shifted slightly from needing 2/3 of the delegates remaining, but not by much. When the dust settles, the bottom line is that they need to flat out win, win, win. And WIN BIG in some states with a decent delegate count. So far, that's MI and WA. In fact, Bernie's got more delegates from the accumulated states he's lost in than he does in the winners.
Momentum? It will mean something going into WI. That's an important Dem state and he's been polling well. WY? Who cares? Truly... if I'm a Dem strategist and considering changing my support to Sanders, I don't give a rat's ass how much he swept Idaho and Montana and Oklahoma by. I care about that baker's dozen of states that the party needs to carry to retain the WH. He's won two of those (MI and WA); he's lost in FL, TX, VA, OH, IL, MA. "Yeah, but he was close!"
Close doesn't matter. Wins in red states don't matter. The Dems have about as much chance of taking Utah and Oklahoma as I have of winning Miss America. When the numbers are finalized at the end of today, Clinton's magic number is going to be around 650, I think. That's the number of delegates she needs to pick up of the remaining 2073, but we'll have to wait to see the final allocation of delegates from today. If it's 660 or 670, it's not much of a difference. It's a very small number.
It seems that you're saying two things about (pledged) delegates.
First of all, that delegates in primary states matter more than those in caucus states, because there is wider voter participation in primaries than in caucuses. (well, you don't outright say that but I it seems implicit in various posts by various posters in this thread).
I think that very much depends on your viewpoint. Do you want a candidate that appeals to the activists (i.e., those who bother to show up at a caucus) or to the general electorate?
Second, that delegates in swing states matter more than those in solid blue or solid red states. Of course it's much more important to know which candidate will do better in, say, Florida, than in Kansas, which will vote red even if the GOP nominates the second coming of Hitler, or in Washington, which will vote blue even if the Democrats nominate LBJ's head-in-a-jar.
But in the end, the rules are: a delegate is a delegate is a delegate. No matter from which state they're coming, and no matter whether elected by caucus or by primary. Them's the rules, and whoever wins by that metric gets the nomination. If the Democratic party puts more value in which candidate is more popular in swing states, they would have given more delegates to Florida, Ohio, etc. But they haven't.
They can change their mind any time they want. A theory is they won't go against the pledged delegates, which would in effect disenfranchise the voters. The we'll see. I still think Bernie's chance of getting a majority on pledged delegates is a longshot BTW.
He certainly didn't diminish his chances this weekend and decreased his trailing by some 65 delegates. But if he wins in pledged delegates, which I agree is a longshot, it will be not a day earlier than California votes.
So will the superdelegates change their mind if that happens? Out of 712, a whopping 469 have expressed support for Hillary against 29 for Bernie. Eight years ago, Hillary was also leading in superdelegates, but by a much more modest margin, and moreover, Obama clinched a (small) lead much earlier in the primaries season so they had ample time to see the writing on the wall. Or will they argue, as above, that primary states are more important, or that swing states are more important and stick to Hillary?