Yeah, I've also found that the abstract approach is what pays the most dividends. When I taught electronics lab, the physics students would get wrapped up in trying to think from the point of view of fields and charges, which honestly isn't that helpful for electronics (unless, as mentioned above, you're doing device physics).

On the other hand, research (by the PhET folks in Colorado I believe) has shown that having students play with a Java applet beforehand (that shows electrons moving through wires) really helps their performance in a basic electronics lab activity. This would be measuring voltage, resistance, current. However it's likely that the sandbox nature of the applet is what helps, not so much the moving charge concept.