Cleveland Ohio had 100% Obama voting in 16 precincts, and more than 93% in at least 100 IIRC. The numbers still don't give Romney a victory, it's about 1200 votes to take from Obama's count and give to Romney. The high 90's votes for Obama are probably more correct than the 93% black exit polling, the latter is probably low since they didn't poll in those 100% Obama ghetto precincts.
As for turnout, generally 60-70% in Obama land, more than Romney territory in many cases.
If all that we needed to be concerned with was the 100% districts, you would be right. But the 100% districts are merely a symptom of voter fraud, manifested in places where there was no need to be cautious, and they overdid it. Voter fraud, however, is not restricted to solidly partisan territory. There are double voters, dead people kept on roles, and lax identification requirements, all of which are a concern everywhere. The 100% districts are an indication of what might be going on statewide. But as I say, nobody on this forum has the resources to thoroughly investigate even Ohio, let alone all of the swing states.
In a landslide victory, such as Obama had in 2008, it’s moot. The voter fraud changes nothing. But in a close election, such as in 1960, it can make a big difference.
Thus, the election might have been stolen.