If you go back, for example, and look at usenet threads from the first few years after Xtol came out, there were reports from experienced users of Xtol sudden death from the 5 liter packets, with the developer mixed with distilled water, properly stored, and using at least 100ml stock per roll of film. If I remember rightly, John Douglas, Richard Knoppow, and John Hicks, among others, had failures. Ryuji Suzuki took it seriously enough to try and solve the problem with a similar developer. I used to argue with J. Douglas all the time. He thought that no one should use it, whereas I liked it. Nonetheless, there was too much anecdotal evidence back then to simply dismiss the failures as a user mistake.
That said, I haven't heard about failures for quite a long time, and so I conclude that if there really was an issue, then it's been fixed.
