I was actually referring more to the original picture where a pullchain switch had been placed through a hole in a toggle switch face plate. I'm sure it worked, but I seriously doubt that the leads on that switch were any larger than #18.
The pullchain probably won't cause an incident, but it would likely work itself loose after several on-off cycles. Then being loose may cause a splice inside the box to break, and so on, to create a hazardous condition.
Oh, you've likely seen wiring done so bad that you could accuse the installer of attempted arson, but this shed switch kludge would be classified as "mildly hazardous". Even if it has odds of 1 to 10,000 of causing an incident in a year, that would still mean that if that kludge existed in each of the, say, 100 million dwellings nationwide, you'd have 10,000 incidents every year. Which is way too many, so the code is written to get that down to just a few incidents a year. Or none at all.
Not to resurrect this thread from the ashes, but as I was reading, I turn my thoughts to the relatively new AFCI devices that are now code mandate. I understand the reason for the desire to have made them a requirement. But in this case is the code not catering to stupid people ? It will prevent that person from burning their house down who doesn't mind their extension cord or their appliance cord that is obviously past it's usable life. I'd like to hear some discussion about that. Because there may be other reasons for their employ, I just may not be aware of it. In relation to the thought of too much safety, as crazy as that thought is.
AFCI devices are not going to object to that Yankee's ingenious 'solution' for there is no arc faulting failure.
The problem with that particular install is that there is no suitable OCPD up the line. Consequently, there'd be no logical reason why one could not overload the conductors, melt through the insulation so quickly that a flame started before even an AFCI tripped.
If the short is SMOOTH and regular, the logic of an AFCI is going to see the cook off as merely a 'hot plate' pure resistance load.
AFCI are going to object to universal motors much more readily -- the commutator -- of course.
One simply can NOT use conductors undersized versus OCPD -- unless the tap rules are obeyed.
Beyond that, as previously commented, the slack installation is not going to stay tight under use. Of all the flaws I've witnessed in installations -- loose EMT, boxes, fittings come at the TOP of the list for hazard. They are the source of rubbed conductors and frayed terminations. When not correctly bonded, they are the source of electrocutions. They happen ALL the time -- typically killing a kid so young as to not recognize that something's gone amiss with the electrical system.
Beyond that, the trade, generally, does NOT favor DIY kludge-ups -- if only for economic reasons. With knowledge comes realization: the DIY crowd is dancing with death -- and dares not know it.
Lousy DIY plumbing attempts reveal themselves as drips and ugly noises. No-one dies.
Lousy DIY electrical installations are not apparent to most -- and the perils include loss of structure and loss of life.
Add to that the commonplace DIY attempt to work on hot circuits! DIY are willing to take chances that no experienced hand would tollerate.