So, how does Hillary win? To the delegate counter! I'm working off the Forbes delegate tool here, so your mileage may vary a bit. It has (after me completing Mississippi for it) Obama up in pledged delegates 1397-1237.
Surprisingly, with one glaring exception (Alabama) the delegate percentage and popular vote share for each candidate in the two-person contests has stayed remarkably close to even, with spillover evenly on both sides, so I'm not going to bother with deep district-level analysis for this.
I'm going to try to be as realistic as possible while giving any subjective range to Clinton. This includes not giving him any gain over time, which he's done everywhere.
PA: Average of the 3 recent Pollster polls is Clinton +16, so that's what I give her
Obama 66, Clinton 92
GU: Everyone seems to split this 2-2, and I have no data to indicate otherwise, so
Obama 2, Clinton 2
NC: Two recent polls are Obama +7 and Obama +8, so I give him the even number
Obama 62, Clinton 53
IN: Only poll we have at all says Obama +15. I dropped it a point.
Obama 41, Clinton 31
WV: Into the realm of speculation here, I average her Ohio result and the PA poll to say Clinton+12.
Obama 12, Clinton 16
OR: Similarities to WA, but a mail-in primary rather than a caucus. Let's go Obama+10
Obama 29, Clinton 23
KY: Similarities to TN. Let's say Clinton+20 for fun.
Obama 20, Clinton 31
PR: Conventional wisdom puts this anywhere from even to moderately Hillary. Let's say Hillary+10.
Obama 25, Clinton 30
MT: Obama+20 is in line with other results.
Obama 10, Clinton 6
SD: Ditto Montana
Obama 9, Clinton 6
Grand Total:
Obama 276, Clinton 290
Add in the pre-PA number:
Obama 1397, Clinton 1237
Final SWAG:
Obama 1673, Clinton 1527
Wow, that's still clearly a win, 150. So... what now? Oh yeah. Superdelegates.
Obama 211, Clinton 247
Super-SWAG:
Obama 1884, Clinton 1774
Still 110. This leaves 337 Supers on the bench, assuming that somehow, Obama stops his trend of getting SDs while Hillary slowly leaks hers. How many does she need for a bare majority? Two thirds, it turns out:
Obama 113, Clinton 224
----------------------
Obama 1997, Clinton 1998
Then she'd need some other as-yet undecided delegates and Edwards votes to hit the 2025. But let's just take her passing him as a win.
Does anyone think that 67% of the Supers will go with the candidate with over 100 fewer delegates and fewer popular votes (this works out to a significant PV loss, still... feel free to do the math, but show your work)?
Of course not. So, this means our rosy-for-Hillary scenario isn't good enough. How much can we bend it? Well, how many supers can she turn? Half means she's caught up anyway. where's the realistic meeting point? How about if she wins 40 more delegates, to bring down the pledged gap to 70 and the overall gap to 30? Then she needs 55% of the supers.
Okay, that sounds like it could theoretically happen. Where does she get 40 more delegates?
Let's say the Indiana poll was way off and she actually wins by 16. There's 11 of the 40. Let's up her win in PR to +20. That's... uhh... 4. And she ties in Oregon. That's 3. And she barely wins North Carolina. There's 6 more. 24 down. And MT and SD are both 51-49 losses. That's 3, total. And West Virginia is a +20 win. That's one more. Okay, add em up... she just needs 13 more.
There's nowhere left but Pennsylvania. What kind of win gives her 13 more?
67%
Even in this perfect world, she needs a 2:1 win, doubling her current lead in the next six weeks to a level she's seen in no contested contest.
I'll say it again: Obama already closed the deal. Just because he did it in Washington, Virginia, and Wisconsin instead of somewhere on the Mark Penn preapproved state list doesn't mean it didn't happen.
It. is. over.