How do you know this? Anyway, you guys are making me wonder. While I expect this industrial product to actually do what it says (it IS a mechanical bond system, that cannot dry out, and has very high sheer strength, in an application with truly little sheer force), still, I don't know anything about it's degradation over time. If it was bad, it shouldn't be in use anymore. But, I could, easily, secure a loop of SS wire around the assembly that would ensure that the little guys could not simply "fall off" and find the sparkingest places available. Anyway, enough about my choices. Back to your regularly scheduled NC400 sonic qualities programming!
I don't know anything about whatever specific glue you used, or much about glue in general, so could be completely off-base. If you trust it, that's fine; my experience is older but contrary (and fairly extensive over a lot of glued-on parts over the years). Part of the problem is the thermal cycling that happens. The metal and glue typically have different expansion coefficients so the glue is repeatedly stressed over time as you turn the amp on and off. That causes cracks that gradually get wider and more numerous, and the glue gets brittle over time from heat and oxidation, so eventually the parts fall apart (usually when you jostle or pick up the component).
Again, glue is not my thing, just relating my experience. Me, I'm from Missouri ("show me") and a skeptic, having been burned by failed glue joints many times over the years (and not just in electronic components; wood glue does not last forever, either). I was also taught to not trust adhesives alone for mechanical integrity over time and stress, but that may be more a function of high-rel systems for which I designed ICs (flight systems, space systems, shipboard and underwater sensors, etc.)