This may build on existing literature. Skin conductance, and startle responses seem more valuable frankly. I'm not sure what this adds. Either way, I'm not sure where the value in predicting political ideology comes from. Maybe we could lesion some folks to do some social engineering?
Third... Yes I meant what I said. I really think we have no idea what regions outside sensory and motor cortical areas are doing. Hell, we don't even fully know what the retina is doing. Anyone who says otherwise is lying. For example, memories are stored all over your brain in every synapse. What would you call the "memory" area? My problem with "middleware" research is that it is all very dependent on the definitions of your terms. For example... What makes an image pleasant as opposed to neutral? People have agreed upon data sets and so forth, but that's just like standard social psychology. Using correlations with old surveys to validate your new surveys. Around and around we go. There are things you can do, but it all depends on how much squish you're comfortable with. I'm not comfortable spending lots of money on lots of squish. Fmri is a bubble that needs to pop. In the last round of cuts they gutted pure behavior (entire study sections at the nih) in favor of few more pretty pictures. That's a travesty. Don't even get me started on the genetic lines of mice for studying "vision." Stupid things can't even see.
Garbage In -- Gospel Out.