The argument that the EC is there to the avoid a deadlock situation is quite inaccurate. In fact, given there are only 538 EC votes to distribute, the likelihood of a tie (269-269) is much, much higher than the likelihood of a tie in the popular vote (where in 2008 both candidates received > 50 million votes). Nate Silver's 538 blog estimates the current chances of a EC "deadlock" at 0.2% -- small, yes, but not so small as to ignore the possibility entirely.
The "solution" under this scenario? If the election doesn't produce a candidate with 270 or more electoral college votes, the race gets decided by the House of Representatives. Can you imagine the reaction if Obama won the popular vote, but only received 269 EC votes and then the (Republican) House awarded the election to Romney?
There is precedent for this, of course. The 1824 election saw Andrew Jackson getting a plurality in both the popular vote and the Electoral College, but not a majority in either. Ultimately, a Congress hostile to Jackson would award to the Presidency to his arch-rival, John Quincy Adams. http://en.wikipedia.org/wiki/United_States_presidential_elec...
In terms of worrying about deadlock, a simple popular vote total is far superior than any permutation of the Electoral College.