The art of making a fugue is to think creatively about how to combine the given material. A musically significant challenge (for the composer and for the performer too) is to recognize within the theme some characteristics which can be elaborated for development through fugal repetitions in the course of the piece. In other words, the observation of musically pertinent affordance in the theme is precisely what makes a fugue interesting. The assumption that, as long as themes and generative rules are decided, the rest takes care of themselves through some probability rules is plain wrong: what divides good and less good compositions, rests on the creative strategies employed in the combinatorial processing of themes and rules as a piece of music. Bach in the above example does not have the most inspiring theme (no attractive gesture nor interval in sight). But he uses its harmonic stability maximally to guide the music to gain tonal variety through imaginative transpositions. Perhaps unusual for a fugue, the performance of this piece may put more emphasis on the harmonic pathway than the theme itself.
The practice of improvisation on a given theme fllows a similar guideline. You play between the temporal/textual development and the thematic character, creating tension and resolution from the potential of the theme. Some experimental music (where basic material and rules are outlined) and what I have called elsewhere as 'prescriptive notation' (notation that gives instructions such as tablature) operate on similar principles in performance in that the tasks of 'sense-making' as music are very much in the domain of the performer.
These observations lead me to posit a hypothesis: understanding affordances in a given situation requires experience and expertise in the field. Knowing all the possibilities in a given situation is one thing, but understanding their affordances is another. This understanding can be gained through detailed analysis, knowledge, experience, intuition, feeling of their presence, or just by trial-and-error. Knowledge and expert skills are equally important in this process.
The ‘unique reality’ found through making most of the affordance in an algorithmic process is real and not abstract. It is mutant, and changes every time the sequence is carried out. But something of the expertise remains throughout like a fingerprint.