Let me focus on just the incentive to cheat aspect. As you noted, in 2012 nineteen states were totally uncompetitive. In those states under the current system there is zero incentive to cheat because the outcome is binary: either you manage to successfully flip a previously uncompetitive state without getting caught or you fail completely. The risk of any kind of cheating at the presidential level in these states simply isn't worth it. This all changes in a National Popular Vote (NPV). Now cheating in uncompetitive states can be very rewarding.
Let's look at a close election: 2000 Bush v. Gore. Under an NPV, Gore's margin of victory in 2000 would have been about 544K votes, or 0.52% of the total popular vote. If you look at just two uncompetitive states, NY and TX*, you need only swing the vote in those states by about 4% to reverse the results of the NPV election (NY: 4%, TX:4.2%), and that's just in two uncompetitive states. Throw California into the mix^ and you now only need to swing the vote by about 2.5% in each state to flip the election. Spread out to all 50 states, you only need to come up with 10K votes in each state to completely reverse the election.
You don't even have to cheat to make this happen. Consider the voter ID laws that are proposed or on the books in many states. There's a reasonable argument to be made that voter ID laws protect the election process by mitigating vote fraud. However, some studies # estimate that voter ID laws depress turn out of lower socio-economic voters, who typically vote for Democrats, by as much as 10%**. If we can assume this is true, and the recently rejected TX voter ID laws were in place in the 2000 election, Democratic voter turn out there may have been lower by about 640K votes; more than enough to flip the election.
Under NPV, all 50 states have a powerful incentive to monkey with their voting laws because with just a little nudge they can affect the outcome of the entire national election. Hence, the eventual outcry for a national system of standards for elections.
* New York: 6.8M votes, 25% margin of victory for Gore; Texas 6.4M votes, 21% marge of victory for Bush
^ California: 10.5M votes, 12% margin of victory for Gore
** State of Texas v Holder http://s3.documentcloud.org/documents/415387/texasopinion.pdf
# 2011 paper by Dr. Michael Alvarez of the California Institute of Technology http://vote.caltech.edu/sites/default/files/vtp_wp57.pdf