The Clinton claim that she has the best chance to win the November election is based in part on her narrow wins in a number of critical states. The argument that since she won a state in the primary she would more likely win in the fall is superficially appealing, but in a case like Florida (where Obama was not on the ballot) the argument is particularly weak. One might also question whether Republicans who crossed over to the Democratic Party in the primaries in order to keep Clinton in the race would be there for her in the fall.
Another question that some might find relevant relates to the integrity of the voting process. Many people in this country have become concerned about whether our votes are being counted properly. While there have always been good reasons to question this process and no doubt there has been some cheating in every election, the introduction of computers to collect and count our votes in secret has made the process even more dubious.
That some precincts are credited with more votes than voters does raise some concern for the process. There is also the oddity that exit polls, always reliable in the past and generally reliable in other countries, have suddenly become unreliable in our own elections.
It seems quite fair to say that we have a faith-based electoral system in this country, particularly where computers are used, and this is especially the case when (as in this country is always the case) the software on these machines is private property that is neither carefully tested nor open to public inspection. Under these conditions, any computer professional will tell you that it is not only possible, but quite easy, to pre-determine the outcome of the vote count.
Some have noted that all of the manufacturers of voting machines have strong ties to the Republican Party and some have suggested that this has been used to give the Republican Party an advantage in recent elections. There are some reports of people observing their votes flipping unexpectedly in favor of Republican candidates, and there are some vote-count pecularities in past elections. I cannot offer any conclusive evidence of a fix, but I don't beleive anyone can prove it has not happened either.
There is some circumstancial evidence suggesting that machines used in recent primaries favored Clinton over Obama in similar ways. For example in New Hampshire, votes counted by hand favored Obama, whereas votes counted by machines favored Clinton. Around the country, caucuses (which do not involve a machine) favored Obama whereas primaries counted by machines seemed to favor Clinton. Again, there is some smoke but no smoking gun and no fire.
If someone is manipulating the vote-count, they are operating under some constraints. They cannot be too obvious in what they are doing and will probably not be able to influence the general election more than they have the primaries. In this connection it is worth noting that overall, Obama seems to be winning in the primaries. To the extent that you can credit the possibility that primaries have been manipulated in favor of Clinton, you should also be able to credit the probability that Obama's prospects for in the general election are better than Clinton's would be.