Improvement is good.
But just comparing the Axe to the Kemper is not an improvement.
Comparing axe to kemper is not itself an improvement, true.
But surely there's been user-made tests that drove improvements on some level. How about aliasing, the illustrations, tests and eventual fix with more processing power allocated to minimize aliasing, even post-fix tests confirming the results?
I like such testing because it helps battle one's own biases. Surely this isn't always welcome. But I also think testing as a whole is elemental for development, on several fronts.
On that end, surely companies themselves do much of this. There would have been no kemper or axe fx without all kinds of extensive testing, never mind a pod. Users can also contribute, even if in seemingly less important ways.
These ways can include reporting back on beta releases, conducting sound tests or presenting other discoveries. Much of this goes on in companies to begin with, of course -- at times, I would argue, on a surprisingly similar level to what some users do.