Interesting thread, however to play devils advocate for a moment, I am quite sceptical that the fuse could be the initial point of failure.
It seems much more likely that the dielectric of the capacitor is breaking down, either due to the capacitors being marginally speced, degrading with age, or a slightly out of the ordinary voltage spike, and that the short circuit current of the failed capacitors would be enough to blow the fuse.
Convince me otherwise.
Another thing that makes me sceptical is that I can't believe the designers wouldn't have performed live tests with a deliberately under rated fuse to observe the behaviour of the unit when the fuse blows, especially if there is anything that could cause a voltage spike when the fuse blew.
Of the reported failures how many are capacitors only, how many are fuse only, and how many are capacitors and fuse ?