Roads states pretty early in the book that this is about his process of composition, and he doesn’t try to cover, for instance, improvisation. Nevertheless, I found some relevance to my process, which often is at least partially improvisational.
He does love to use granular synthesis/microsound as the example for everything, and his favorite word “multiscale” turns up a lot. But it’s a pretty useful way of thinking.
I guess my main criticism is that he tries a little too hard to make the case that electronic music freed composers from constraints that they previously felt.
Pre-electronic Western composers did have a tendency to treat notes as indivisible, homogeneous, atomic units; 12TET tuning mostly dominated to the point people forgot other alternatives existed, and timbre and spatialization options were more limited. But those were not absolutes, particularly for performers – it was understood that composers wrote notation, and conductors and performers interpreted it, adding expression. (A few composers even insisted their written scores were the music itself, and it didn’t matter whether anyone heard or performed it.)
Those same biases were maintained in electronic music in the form of keyboards, notation software, the design of MIDI, piano roll sequencing, etc. In fact what MIDI calls expression is anything that deviates from notes that you can play on a piano.
I found it curious that Roads didn’t associate his argument with the point that, in electronic music, there is usually no division between “composer” and “performer.”
I do appreciate his thoughts that formalism and determinism really only take you so far, and that algorithmic composition processes are weakest where it comes to matters of form – that music needs human direction and insight at crucial moments.
Anyway, it was a thoughtful and thought-provoking book, but as usual, I want to read something lightweight and fun now. 
