It's funny that so many people don't understand this failure mechanism in tantalum caps. This is not about voltage derating. This is not about surge currents. This is about a failure in the dielectric caused by the high temperatures of reflow soldering. They have the same problem in manufacturing, but they properly deal with it in test.
The exactly problem is that the microscopic shorts created, cause localized heating around the flaw. This causes a chemical reaction that insulates the short, but also a release of O2 which can allow materials to combust, raising the temperature higher. Clearly, this can get out of control easily if the initial heating is very severe.
On initial application of power to a tantalum cap after an event like reflow, there needs to be a current limited application of power, (much more significant that just the surge current ratings) to prevent the localized heating from becoming severe enough to cause release of O2. Once the capacitor is powered up this way, the small shorts are isolated and the problem does not need to be addressed again.
I can see why this would be confused with a surge current concern. I can also see how people might confuse this with the derating factor to extend the life span of the device in use. But some people are resistant to the idea this is a separate issue entirely. I've never seen a name for it. That might help a lot.