I've mentioned the analogy dozens of times. Wheel bolts or studs are not in shear. It's the friction between hub and wheel that carries the vertical load and transmits torque. Similarly if the QR plate does not have a compliant surface and there is metal to metal contact to the scope foot then there is no shear on the bolt or anti-rotation pin.Agreed about the spokes, but you're talking about (almost) vertical loads (compression/tension). If the 'scope is always horizontal on a tripod, there's no problem but the turning moment on the 'scope/QR plate interface is appreciable for a 2 kg 'scope, when it is being carried in an other-than-horizontal position and being bounced about.
A friend of mine, who is an authorised dealer for all the major manufacturer's products, says that the majority of 'scope failures are based around the mounting system. I would still prefer a more robust material for the tripod attachment but I agree - this is the area which lets pretty well all 'scopes down.
When you consider the hardware used to attach a small camera or a Nikon ED50 to a tripod is also used by many people for a Swaro BTX115 is a nonsense. In the latter case, to rely on a 3/8-to-1/4 adapter with only 2-3 threads making contact is rediculous. The blind hole in the foot should ideally be about 15-20 mm deep. You are right - the manufacturers of optics are partly to blame but they arn't helped by the makers of tripod heads!
If everything were engineered optimally with sufficient thread depth then a 3/8" UNC bolt could sustain about 40 N.m.
Tighten it to a mere 5 N.m and you still have a compressive load of more than 2500 N or 250 kgf and prudent lubrication of the thread could only increase this. No Loctite required!
A blow with a hammer is not to be recommended but it would be unlikely to loosen the thread connection.
John