But a duplicate of the actuator failed in a lab on Earth after almost exactly as many turns as the actuator in space lasted -- that's how they determined the cause of the spacecraft's going offline.

True, but that confirmed the test run in space, not the other way around. It may have be hard to know in advance if 348 failures on Earth is representative of space, especially because it wouldn’t just be that one system. It would have been lots of parts built in to system working together.

Original mission for Voyager 2 was Jupiter and Saturn fly-bys which was mostly a success. The idea is to engineer parts to survive the mission and maybe go a bit further if you aren't at a point of diminishing returns in regards to cost. If it costs $1000 for a 10% gain in life of part, maybe go for it. $1M for 10%? Maybe not. Additionally, gamma radiation can do some weird things to materials and can effect things differently depending on dose rate. Lubricants can increase in viscosity and acidity under gamma. Radiation effects on spacecraft was not as big of a deal back then. Missions were short and electronics weren't small enough to have noticeable radiation effects.