> When electronics fail, it's generally a handful of causes, or one or two well-known design flaws in some particular model of device.
Once upon a time in the age before I dedicated my life to all things software repairing color tvs was one my sources of income. I respectfully disagree with your assertion, electronics can fail in very many different ways. All the way from 'rodent electrocuted in HV power supply' to passive component that never fails somehow in fact did fail.
Even something as mundane as a coil-over-a-resistor can fail if the conditions are right. And don't get me started on bad soldering connections, corroded connectors and a million other little details that can affect the functioning of a device to the point where the repairguy gets called.
This was before the days when oscilloscopes were portable, and unfortunately also before the days when I had a driving license. A typical house call would end up with a working TV or VCR and me learning yet another way in which things could fail.
Main tools: a couple of folders of schematics of the most commonly encountered models of TV/VCR (pick the right one before you leave, hope the user read the type plate correctly), a stack of matchboxes with passive components, some of the more common active components with enough spread that you could cover most cases, a soldering iron and a very large collection of parts scavenged from old TVs waiting for the opportunity to recycle them.