That he had to labor to achieve and improve his ideas and visions, and that he did so constantly until the very last day of his life is obvious. But experimentation for him was only a mean to an end, it was never an end in itself. Is my argument really that hard to comprehend?
You just reframed your argument, contradicting your previous assertion that Bach was never experimental.
But, taking your new argument for what it is--because everyone deserves the opportunity to say what they mean and sense--I still don't see how you can claim to know how he worked, without assuming that he exhibited some blanket cultural functionalism. Excluding the church cantatas and passions, and organ chorales, etc, he wrote secular music.
The WTC, which is stylistic and contrapuntal exploration, and perhaps investigation and testing in all keys, major and minor,
twice (two books, the second published twenty years later). Why all keys, why twice? Why the first book's "well-tempered" title? We don't know the answers to these questions, but the WTC seems to fit at least a couple definitions of experimentation. One could also say that his trying it once and then again suggests some and further discoveries were involved. Actually he reused some material from Klavierbüchlein für Wilhelm Friedemann, so one could say he did it three times.
The Art of Fugue is clearly an exception, even if nothing else is. It is practically an experiment by definition; a variable, or theme, is manipulated along with its fugal context, and the results recorded.
The work seems to be unfinished, and we don't know the circumstances of why or how. But, it has been argued that the tribulations are right there, and you also have different versions of the Johannes Passion, the 2nd Orchestral Suite, Harpischord concertos, etc, etc. Unless you can communicate with the dead, you can't make assumptions about how his "vision" was realized, and he could be lying to you anyway.