Electoral College Reform: I think Kaus is right that the proposed initiative to divvy up California's electoral votes is much ado about nothing, but I would like to address the unstated assumption that this sort of thing would be fair and appropriate if done nationwide, as opposed to targeting specific states. The biggest problem with the Electoral College is the imbalance in favor of low-population states, like the Dakotas or Wyoming, which enable the voters in those states to have a disproportionate say in the outcome of the election. The other big problem, which partially counterbalances the first, is that a candidate wins all of the electoral votes in a state, no matter how close his margin of victory.
The proposed "reform" on the June ballot cures the second problem, but exacerbates the first. Kaus correctly points out that gerrymandering has made most House seats non-competitive; in California, the lines were drawn pursuant to a gentlemen's agreement in 2001 to protect both parties' incumbents, freezing into place a significant partisan advantage for the Democrats incurred from the previous redistricting, which was done by a panel of judges. At the time, the Democrats thought they were preserving their party's advantage into the future, not believing that it could be conceivable that they would be able to expand their advantage.
As it turns out, the post-Prop. 187 shift toward the Democrats in California was only just beginning in 2000. Only the incumbent-friendly lines drawn in 2001 have kept the Republican Party from disappearing into marginalization in the Golden State. Thus, the lines in California favor the GOP more than their actual strength would merit, and in a Presidential election, it would take a landslide of historic proportions to give the Democrats more than a fifteen-seat edge in the Electoral College under the proposed reform.
But in a state like Florida or Ohio, where the lines were rejiggered to maximize the dominance of one party, even a close election could result in a lopsided electoral count. It is almost certain that the GOP would win a majority of electors in a close election in Florida, even if Hillary or Barack were to win the state, simply because of how the House districts were drawn last time to favor that party.* It is certainly conceivable that a similar scenario could happen in Ohio, a state where a number of troubled incumbent House members were able to win reelection in 2006 in spite of the toxic political situation for Republicans in that state.
In any event, reducing the influence of the large states by dividing up their electoral votes only increases the power of smaller, more homogenous states. That doesn't seem like much of a reform to me.
*The GOP has a 16-9 edge in the state delegation, in spite of it being the state that perhaps best exemplifies the even divisions in the country following the 2000 election.