More prosaically, the 2N3055 has been around so long that it has been 'reimagined' in what must now be 10 generations of semiconductor manufacturing processes. Hell, there probably isn't a fab in existence that's capable of making a genuine original spec 2N3055 in the process it was originally made in. The upshot of this is that if you want to know the actual parameters of a part labelled "2N3055" you need the specific data sheet for the specific variant of the generic 2N3055.
It should come as no surprise then to find a generic 2N3055 emitter base reverse breakdown of 7V and an actual one for the part on hand of more than twice as much. That's way too much of a difference to attribute to "engineer's safety margin fudge factor" but isn't too much for "The actual process for this part is way better than the 2N3055 spec calls for, but we'll just specify the official 2N3055 figures so people don't point to the higher specified figure and think it implies that the part is in some way incompatible.".
It's worse than that. A lot of "legacy" transistor types being listed by manufacturers don't even try to match the original silicon. They just put a die from a "better" spec current process. They just list the headline specifications and don't worry about the detail stuff. One thing that stands out is the physical size of the die. They shrink over the years.
I once emailed the maker of a 400Hz 3ph bench supply I was tying to repair (Behlman IIRC) and ended up conversing with the designer. I received this comment:
"Be careful if you have to replace the output transistors. The real 2N5259 is a very special transistor. Originally made by RCA, then Hitatchi, then SG. It not only had a great operating spec, but the gains were well matched without having to do gain matching, and, most important, the Safe Operating Area (SOE) was quite wide. We had a SOE specification of 1 A at 140 VDC. We had to test each transistor to make sure that it would meet this spec. When the plant was moved to NY in 1990, the new production manager thought individual SOE tests were too expensive, so they were stopped.
In 1994 we started having trouble with the amplifiers operating properly and or destroying themselves. It turns out that SG stopped making the 2N6259 using the original RCA die masks. For whatever reason, they decided to take another transistor die that had the same DC specifications, gain, power, voltage, etc. and marked that as a 2N6259. When we tried to use them in our amplifier they had stability problems, because of gain issues, both in matching and wider bandwidth, and kept killing themselves for no reason, which was traced to a poor SOE, which we were not testing anymore.
We also found out that you could still order 2N6259 marked transistors from other vendors, like Czech or Polish vendors, but none that we tested used the original masks so they did not operate properly, and definitely could not be mixed with the good transistors in the same amplifier. We had to redesign the amplifier and had to re-trim the Driver for new transistors altogether.
I've periodically tried to find replacement transistors for the 2N6259 without success. I also see a lot of units come back with dead replacement transistors in them.
If you have to replace the output transistors and you cannot get a 2N6259 made before 1994, then you take the chance of the amplifier killing itself frequently, depending on the load and usage. You also run the risk of the Driver Card not being trimmed for the gain properties of the new transistors and the amplifier being prone to oscillation. Even one replacement transistor in a 100VA amplifier is a 25% change, so it can be significant."
So even the professionals get caught out.