If kept in continuous use, the MTBF (Mean Time Between Failure) is typically 200,000 - 600,000 hours, depending on size (200k at the smaller end, 600k at the larger end). Whether or not you accept calculated MTBF values as a legitimate indication of equipment life is a debatable issue, but it's about as close to a scientific evaluation as you can find. Other than that, everything is anecdotal. I am routinely called upon to replace 20+ year old VFDs that are still working fine, but the users have a similar fear of what happens WHEN they fail, because in those cases, the components to REPAIR the drives are no longer made. That actually becomes a more salient issue and as a general rule, I recommend people start considering upgrades at around 10 years, depending on the cost of unexpected down time. So for example if you have built-in redundancy in your system, having to wait 2 weeks to repair a drive may mean you can keep them in service for longer, i.e. 20+ years. But if your system would go completely down if the drive fails and there is no replacement available, taking an extra day to figure out how to set up and interface to a new drive might cost you a LOT more in down time than it will to keep your drives current.
" We are all here on earth to help others; what on earth the others are here for I don't know." -- W. H. Auden