I think the argument is that winning a large city such as Chicago, Miami, Atlanta could no longer be used to win high value states.
I'm not sure I buy the argument. Like you say, someone would have to run the numbers for the past several elections. However, even those results would have to be taken with a grain of salt because how an election turns out depends on how candidates campaign which in turns depends on the rules. For example, in 2000 with the current rules it made no sense for Bush to spend much time campaigning in Texas but under a popularist system it would make a lot of sense(*). Thus it is kind of silly try to use the popular vote results from 2000 to predict who would have won under a popularist system.
(*) Which is one of the arguments against a popular vote for president. Candidates pander to voters that give them the most votes per unit effort (e.g. advertising dollars, candidate time, etc.). While your votes might count equally in a popularist election, the amount of effort to win your vote varies widely depending on geographic location. Thus a popularist election would not give voters equal political power.