The electoral college has several distortions. The big one and it's raison d'etre is that it distributes power towards rural states and away from population centers. There's little question which party that favors.
Now it's certainly possible that Romney could have won the popular vote and lost the electoral college due to the first past the post distortion that dilutes big majorities. But that really says more about the relative competence of the two campaigns then it does about whether the electoral college structurally favors the Democrats. Romney clearly needed to run quite a bit farther to the middle than he did.
One way to evaluate claims like these is by looking at an extreme comparison:
Alternate World 1: red states are split into mini-versions of themselves the size of a single CD, a more than doubling of their electoral value. Ignoring effects on the Senate of course, this is just a test of POTUS electoral procedure.
Alternate World 2: 25 very red and 25 very blue states of combined identical population are subject to the whims of a single swing district that is right in the middle of the spectrum.
It should be obvious that the sorts of presidents you get in world 2 look a lot more like those in popular vote world than the ones you'd get in world 1. Actually we might expect that the liberal candidates in world 1 are nearly to the right of the conservative candidates in world 2.