I've found for me, it depends what I'm trouble-shooting. When I go into a system I've never worked in before (and I have the leasure to do this) I like to sit and break it down in my mind until I understand the complete operation; not just the part that is failing that I have to fix. That way when I go back in the future, it takes no time at all to diagnose a problem, because I'm already familiar with it. But this is all electromechanical; even with the schematics, it's pretty linear.
I have tried and tried to learn solid-state trouble-shooting, but I can't visualize the operation of the circuit in even the simplest of applications. And maybe that's a key hangup, what I'm trying to do is see the operation in my mind, in this case, no moving parts I'm actually thinking about current flow.
Maybe I need to find someone experienced who can actually teach it, but I think that some of these skills also depend a lot on the way the learner approaches the problem. Solid-state seems like more of an abstraction, and is accordingly much harder fo rme to grasp (surprise, surprise, I don't intuit algebra, either).
But I definitely think there are definitely different types of trouble-shooting, and depending on the person, some can be more easily taught than others.
My $0.02
-John
