Basically, according to the Caroline Products Theory, the Supreme Court will only step-in to change a result of the democratic process if there is a failure in the process. For example, if you can make a strong argument that the situation with FL and MI is the result of a FAILURE in the democratic process, you get a strong argument for the court (or here, superdelegates) to step in and do something different from what the democratic process has produced.

The reason the Supreme Court has this theory is to avoid being antidemocratic or countermajoritarian. See, we have a "democratic" political process that determines "winners" and "looser." But our constitution gives the Supreme Court (or one can argue the late Chief Justice Marshall claimed) the right to review results of this democratic process: laws, legislation, etc.

The problem for the Supreme Court is in deciding when to exercise this power? The Supreme Court has seen wise to say that they should only use this power when there is a flaw in the democratic process. This way, the court is not imposing its views on the rest of the country but is instead correcting flaws in the democratic process to allow democracy to function better. It corrects the process when it misfires and avoids a countermajoritarian result.

Of course, there are differences between the Supreme Court and superdelegates. For one, an argument can be made that, unlike the Supreme Court, superdelegates are a part of the democratic process. However, because each superdelegate has a vote worth over 10,000 times my vote, they look more like the Supreme Court, where each justice has a vote worth over 27 million times my vote. But there's also the fact that the superdelegates voting IS a part of the "democratic" process in the sense that they have to vote before the nominee is decided, whereas Supreme Court decisions are seen more as being something that comes after the bargains have been made to undo or reinforce whatever bargains were struck by the different interest groups. So in that sense, one can argue that superdelegates are, indeed different from the supreme court.

However, this difference notwithstanding, I think the practical effect of superdelegates overturning the popular vote and the Supreme Court overturning legislation is unlikely to be different. So the superdelegates, in making their decision, should look to what the Supreme Court has done as a model for how to make their decision. If they should decide to overturn the popular vote, the superdelegates should base their decision on correcting structural problems in the nomination process, not on substantive believe that they know better than the rest of us. That would avoid the countermajoritarian problem inherent in have a few overturn the will of the majority--even if it should be a problem because they have the express authority to do so.