MMC Testing vs equations (plus MMC bulk update)

["acmnovak" <acmnovak-at-msn-dot-com> (by way of Terry Fritz <twftesla-at-uswest-dot-net>)]

Greetings,
Earlier today, I went through Terry's MMC power handling calculations. I made a
few mistakes at first, but I think I got it right... finally. The capacitors
are the ones in the latest bulk buy. They are rated 2000vdc (700vac) .068uF
with a lead spacing of 27.5mm... I plan on using 13 in each string,
Do these results look typical of a well made mmc?
Ecap=.08872 Joules Per Bang
Rcap=.02029 Ohms Internal Resistance
Wo=.07152 Watts Per Capacitor
To=2.86085 degrees celcius
Life Expectancy of ~555 Hours
Does that all add up?
 
Wouldn't the bleeder resistors play a role in any of these equations? 
How do you know how long the cap will last? I'm sure I can't time it down to
the hour. I usually keep runtime down to 1-2 mins at most. I don't know how
anyone could kill a cap by using it for 555 hours. BTW- Is that time
continuous, or intermittent operation?
 
Also, I tested a cap to destruction earlier today. It took about 1 minute of
run time at 4kV firing voltage to kill it.  Before that, I ran it at 2kV for
3-5 minutes with no heating....Using Terry's equations, shouldn't it last over
40 hours at 2kVac to kill it?  
Upon autopsey, I found that every layer had been punctured. There were two
dissimilar burned holes all the way through the cap on opposite sides... Ouch. 
When it failed, all I heard was a quiet little snap. No cracks in the
potting...
Don't failed MMC caps usually just explode?
 
For those of you still waiting by your mailboxes for these caps...
I called the distributer yesterday, he told me the caps would be in the mail
today. Aparently <http://www.illcap-dot-com>www.illcap-dot-com  has quite a good
quality
control team. They have to check each of the 500 caps I ordered. : O 
The digi-key stuff is coming tomorrow. So, if all goes according to plan, I'll
have your caps in the mail by monday.