> Id rather use rational techniques to track the outcome of individual policy proposals against some sensible definition of better that could come from a rough consensus.
That's the problem, the definition of "better" is the whole point.
Say for USA, setting the marginal tax rate of 1M dollars to 100%, including for company revenues, and redistributing that money to the poor greatly increases the Gini coefficient. That's just increasing a number that is widely agreed to be a "sensible definition of better", can't get more "rational" than that.
You can write a bulletproof proposal that can't be rationally argued against.
Ideology is not just a series of individual policies, it's a lens through which you can evaluate every possible policy proposal.