I think that due to Internet and all this capacitor -thing has been blown out of proportions.
Yes, it is an undeniable fact that electrolytic caps do have a service and shelf life. However, the practical lifetime depends on overall construction of the capacitor and the operating conditions; one surely will die faster when operating under higher temperature and ripple currents. I've seen the "15 - 20 years" -figure for overall estimation quoted for few decades so I have strong doubt that the figure might no longer apply to modern components, and keep in mind, we have examples of entirely new capacitors that for some reason failed within the first year of their service life.
At best it's an estimation. ....with tolerance of +/- 30 years...
I sure can't generate any kind of statistic. From the various stuff I fixed only few old amp had issues with aged capacitors, and issues with age-induced corrosion, wear, intermittency and component value drift seemed to be far more popular than issues with aged caps. I'd put aged capacitors far down in the list of usual causes for amp issues. Never had any issues with old caps exploding or simply working inadequately without "reforming" them (et cetera) either.
It's pretty telling that we only need to look at this thread and we already have folks blaming old caps for the hum that is actually produced by their pickups, or we have comments about value drifts in other ordinary components too, not just electrolytic caps. As said: blown out of proportions.
Yes, the issue is real but generally I would only worry about it when it indeed becomes an issue. It might take only a year, or it might take fifty years. There is no universal rule of how fast a cpacitor fails due to its age as far as I know.