I'm a little surprised to see this thread resurrected. A couple comments:
Yes, the capacitors I dissected were bad quality; that's the point. They were the kind of thing people were buying and having problems with, so those are the ones I looked at. The point was to try to figure out why these seemed to be failing. They clearly failed because of a bad internal ground, something I wouldn't have expected. So, I'd say that these did indeed fail because of poor design. Of course, a bad external ground could cause problems too, but that's not a capacitor failure.
Also, regarding, "overheating because of a bad ground causes failure." What kind of failure OF THE CAPACITOR, resulting from overheating, are we talking about? In the capacitors I examined, the capacitor puck itself hadn't failed, it just lost good contact with the inside of the can because of burned contacts.
Finally, to clear up a point--the capacitance (microfarads) of a capacitor is either right or wrong at the outset; it won't change in use. Its charging rate depends on its current ONLY--charge is the integral of current over time. The rate of rise of the voltage, if that's what you mean by charging rate, is the current divided by the capacitance. It's not an inherent property of the capacitor. The only way this will change is if there is some significant resistance in series with it, such as caused by burned internal contacts. Maybe that's what you're measuring when you check rate of charge--presumably you're using some kind of standard tester?
I'm glad there is a source of good ones, but apparently most people, myself included, are not aware of this, or know where to get them. I think that the option of using a general-purpose electronic capacitor is a perfectly valid one, as it totally avoids the problems I identified, is easy to obtain, and will last virtually forever.