Our tolerance to training volume is certainly a function of training experience. The term training age is sometimes used to describe the total number of years over which an athlete has trained, particularly in the context of younger athletes. However, this term tends to oversimplify the problem, since our training can be contiguous or intermittent, and can vary greatly in quality and quantity.
Nevertheless, it is well understood that our ability to tolerate exercise is a function of our level of training, recovery techniques, diet, quality of sleep, and genetics. And contrary to popular belief, especially amongst amateur enthusiasts, there is an undeniable practical limit to the amount of training volume we are able to tolerate, governed by our hormones and the rate at which biological and chemical processes are able to occur.
On a biological chemical level, recovery from exercise has been studied extensively. Indeed, the processes of stimulus, fatigue, and recovery were described by Russian physiologist Georgy Vladimirovich Folbort as early as 1941, expanding on the early work of Hungarian-Canadian endocrinologist Hans Selye. But on a broader macroscopic level, it is almost impossible to formulate a research question that would describe the absolute limitations of the rate of recovery, due to the number of variables.
The quantity of work which we can do and recover from is known as our maximum recoverable volume (MRV), and a prediction of that volume is instead derived from historical data from all athletes, and the pattern of progress from the individual. That is, we predict what an individual athlete will likely be able to tolerate based upon historical data from athletes of similar age, sex, somatotype, and level of training, and we then monitor the individual's training and recovery, and adjust it accordingly.
“Therefore there is a golden number of sets one can do to achieve maximum muscle growth, and more than that number is utterly stupid, dangerous or entirely useless.”
What you are describing is probably more due to a well-established gym culture than it is supported by any concrete, authoritative research. The classic 3 × 8–12 hypertrophy training regimen, for example, is only loosely supported by science, and fails to account for other well-established bodybuilding norms like drop- (descending) sets, as well as the observation that track-and-field athletes, cyclists, and athletes from a host of other disciplines experience impressive hypertrophy performing hundreds or thousands of repetitions. So it should be unsurprising that the same thing applies to common notions of the "correct" training volume.
So yes, our maximum recoverable volume is trainable, as well as highly individual. But whilst science is helping us push ourselves gradually further than ever before, it is very much subject to the law of diminishing returns. There are practical limits, and our training volume is only productive if we are able to recover from it fully.
I hope that is helpful.