If you haven't read the Statesman's two-parter on the decline of mixed-politics communities (Part One, Part Two), I recommend checking them out. A couple of my own thoughts:

As author Bill Bishop noted in Lasso, the numbers he plays around with are those of voters, not overall population. It would be interesting to get an idea what if any correlation exists between the proliferation of so-called "landslide counties" and voter turnout. Do competitive counties have higher participation rates, or does an increase in single-party identification drive people to the ballot box? Also, how do turnout rates compare over time for officially nonpartisan elections in places that have them (such as Houston's municipal elections) and regular Dem versus GOP elections?

Jerome Armstrong has noted that there has been a significant increase lately in people who respond "other/not sure" when asked what their party affiliation is, which leads him to speculate that the opening for a real third party is getting wider. We already know that in Texas at least, third parties like the Libertarians and Greens get much of their overall vote total in elections where the candidate of one major party runs unopposed by the other. Are there any national trends here? Do the Greens/Libs/Reforms/etc do better overall in landslide counties?

Finally, as this Chron article notes, of the 38 counties with more than one million people in them, four of the fastest growing ones are Texas counties (Tarrant, Harris, Bexar, and Dallas, in that order). All but Tarrant have been trending more Democratic, which is good news for Democrats overall. (Already very Democratic Travis County, which has about 850,000 people, grew at a 5.5% clip during this time, which puts it a smidge behind Harris and Bexar in growth rate.) Balancing that out is the fact that the six fastest growing counties in Texas with at least 100,000 people are all Republican strongholds (Williamsom, Collin, Fort Bend, Denton, Montgomery, Hays). Don't expect the trends Bishop is talking about to change any time soon around here.

(Note: the Census data from which that latter Chron story came is here.)