Obviously, any time you generalize a party into two vectors like this, it's not exact, and I'm sure there are numerous cases of policies from both parties that don't fit this nice little mold, but I think that is what the OP was trying to point out. The Economist's definition of 'liberal' does not mean 'democrat' as it does in most of America today. They espouse 'classical liberalism', which is more libertarian in nature, but not taken to the extreme. (They, for instance, support the idea of central banks and fiat money, and a carbon tax or cap and trade scheme.)
I'm not really old enough to remember, but my understanding is that the Republican party used to be more aligned to these views before they embraced the religious right to win elections.
government bailouts, payroll taxes, immigration law, sales taxes, teachers unions, Social Security, eminent domain, death taxes, stadium deals, cab medallions, government pensions, whatever it is that the NSA does, Medicare, Amtrak, income taxes, city planning commissions, stock price lawsuits, mandatory government fees, property taxes, random police searches, wealth taxes, rent control, energy taxes, airport security, what the CIA does to people, mileage taxes, corrupt politicians with dirty money in the freezer keeping their jobs, sin taxes, wiretapping backbones, tax credits, alternative minimum tax, airport landing slots
My point is that there is a revealed preference at work here
But the debate reminds me of this post: http://lesswrong.com/lw/jb/applause_lights/