Sound is subjective. A user should satisfy the ears first. Whether that takes EL34 tubes, high-speed ring emitter transistors, capacitor output coupling, direct intestage coupling, or a handful of cat hair in the tweeters, only the listener can answer that question.
Reliability is a different matter altogether. Instead of an opinion, I can relate only observations.
Sansui had decent tube equipment (their roots). When semiconductors became prevalent, it apparently confounded their engineers. Everything built from the 3000A to the G-9000 displays a tendency to operate far too close to the edge of safe operation. We all know the manifestations of that, with frequent and inexplicable cascade failures in driver circuits, protection circuits which don't, and some truly Goldberg-inspired factory modifications intended to address problems. I suspect it was a function of market pressures, and trying to get items on the shelves before the other guy with less regard for durability testing. Some of that can be re-engineered out of the equipment based upon our years of observing the failures and repairing, and some cannot.
Pioneer seemed to have a handle on the semiconductor realm, and typically designed into their unit more than adequate safety factor, overhead, and reserve capacity to remain well within margins of safety. But somehow they still managed to consistently design and locate many of their power supplies in the hottest, least ventilated ares of their boxes. We've all seen how that plays out as well, and have learned the common remedies.
The short version is that I listen to a lot of Pioneers stored in racks while working on Sansuis which hit the bench.