Who Benefits from National Popular Vote?

This Res Publica & Cetera post got me wondering about something. We typically assume that the Electoral College disproportionately benefits small-state voters because of their disproportionate representation in the Senate. What I'm wondering is: doesn't it also disproportionately benefit voters in states with large non-citizen, felon and under-18 populations, who are disproportionately represented in the House (since House seats are apportioned by total population, not by number of voters)? I assume that the first effect is more substantial than the second, but it might not be that much larger.

What I'm trying to get at is that, while it's typically assumed that switching from the Electoral College would be a boon for big states and a loser for small states, this isn't necessarily the case. The biggest winners may in fact be voters in medium-sized states with small immigrant populations, like perhaps Kentucky and Indiana.