If you are looking for some good news to cheer you up, Obama is quietly winning his add-on superdelegates while Clinton is not. The superdelegates are obviously still greatly aware of who the nominee will be.
In case you don't know, add-on superdelegates are those delegates that are technically superdelegates, but are chosen AFTER the states have voted. There's a good description here. In summary, each state selects a few "bonus" delegates after it's primary/caucus. The selection process varies, some using State Committee nomination while others using pledged delegate voting. There are 76 add-on superdelegates, and 17 have been selected already.
It has generally been assumed by many people who model delegate projections that whoever wins the state will win the add-on superdelegates. While a reasonably good assumption, it has not exactly held so far. Using the "state winner gets the add-ons" model, Obama should be losing the 17 already selected add-on delegates by a count of 8-9.
However, only 12 of the 17 add-on superdelegates actually endorsed after being named, and Obama is winning them 8-4. He has gotten 7 of the 8 delegates he should have by the model (1 undecided in Missouri), and he has "taken" 1 of Clinton's add-ons (from Oklahoma). Clinton has only 4 of the 9 add-on endorsements that she should have by the model. She "lost" one to Obama in Oklahoma, and 4 of hers are still listed as undecided (NM, AZ, and 2 in TN).
In summary, Obama hasn't had any "defectors" and only 1 add-on has failed to endorse him (yet). Clinton has had 1 defector, and HALF of the 8 others haven't committed to her. Obama is batting 87% (or even 100% if you count the "steal" in OK) and Clinton is only batting 44%.
This bodes well for Obama. Many of the delegate projections you see (e.g. kos, Chris Bowers) assume the "state winner" model for their add-on superdelegate totals. Bowers has noticed the model failing, and has hinted about dropping it. The model only "fails" from Clinton's perspective, and is likely to lead to an even bigger delegate margin for Obama than has been projected so far.
FYI: the model predicts that of the states that have voted so far, Obama wins 37-31 in add-ons (although as pointed out he's beating that projection so far). The still upcoming primaries would be expected to go 5-3 Obama, with whomever wins Indiana getting an additional one. With Clinton needing the vast majority of the remaining supers, the add-on process is really hurting her remote chances.
Obama also has another likely positive deviation from the projections. The model assumes that the California, New York, and Texas add-on delegates will go to Clinton 12-0. Obama can quite possibly be expected to take 1 or more add-ons from each of those states.
--
Note - more good news ahead. In the next 6 days before NC and IN vote, 11 add-ons are up for selection: 2 in MD, 4 in NY, 1 in LA, 1 in SC, 3 in IL. By the "state winner" model, Obama should outgain Clinton 7-4. It's not likely that he loses any of the add-ons from MD, LA, SC, or IL, and he might take 1 from NY for an 8-3 split.
No matter what, it's likely that Obama supporters will have good news on the add-on superdelegate front from here to the NC/IN primaries. Add-ons will likely erase half of the 10 delegate gain Clinton got from Pennsylvania. Perhaps other superdelegate announcements (like Bingaman today) will erase the other half. We could end up starting the day in NC/IN at the same place we were before Pennsylvania.