Shutter can fail for many reasons, certainly speed is a big factor. Using one at 1/8000 sec all the time will cause it to fail sooner with all other things being equal.
Then lucky me Canon has disabled 1/8000 on the 6D (though you can set it with ml)... maybe they figured with the shutter quality they put into the 6D, using a speed that fast would let them drown in warranty cases?
I really hope the shutter in the 6D is as sturdy as in my 60D which now has 150k cycles... but I doubt so since the ff design causes more component stress
Its hard to tell if the 6D shutter speed is a feature being limited for product differentiation, or because its not a robust design. In either event, unless a shutter is fired at max speed all the time, its going to last a long time on the average for a normal user. The only shutter that has failed for me had a thumb put thru it on purpose. I was using it for the Canon CLP and I wanted to see what it took to damage a shutter on one of the old 35mm film bodies. (It was a T50 as I recall), and already had a malfunction. The shutter was actually pretty strong, but once damaged, it could not be fixed, it had vertical metal blinds that bent and could not be flattened out again.
Its statistical, one 60D might last for a million actuations, another might fail at 100. Unless you test 1000 cameras for 1 million actuations, you won't really know how good the design is. Most life tests elevate the temperature and then extrapolate the life based on how many actuations it takes to fail at a very high temperature, probably at least 100C. The test temperature depends on what materials were used, you can't use a temperature that causes material failure before mechanical failure. That means its just a prediction based on theory. Changing assumptions a little can let you predict most any life you want, so the result has to be compared with real life to validate it. A formula that correctly predicted shutter life for a old model is used as the basis for a new one.
Many insulation materials and lubricants can handle 125-150C with no problem, same for IC's and other electronic components. At very low temperatures, stiff lubricants, and contraction of the parts means very tight joints and high friction that puts extreme stresses on all the mechanical parts.
Virtually every failure is mechanical in nature. There are a few where there is some type of voltage breakdown in a semi-conductor, but mechanical breakage is much more common. In IC's, for example, its often a internal lead or contaminant that causes the breakdown, these are mechanical in nature as opposed to the semi-conductor failing.