So note that the I feel like my definition of stand-alone instrument would be something like “a system that can create or alter audio in ways that are interesting, without necessarily relying on things outside of itself (but it could), and should aid in getting something interesting without causing creative fatigue”. This is extremely broad (and probably a little against a lot of the stuff in this thread), which is why I wanted to say it up front. Another thing I think what makes the system an instrument (or rather “good” at being an instrument or not) is pulling in these user experience points like you’re mentioning in your post, @papernoise. And I just want to point out that I feel like it can be a lot easier to try to explore and plan using these sort of aspects to define what I’m looking for, rather than focusing on the “sound”, because ultimately, I got into (and I think, generally, people get into) modular more often than not to try to explore and experiment with sound.
This is something I’ve been focused on recently. I put away my BSP and SQ-1 in the closet and have been focused on using the teletype exclusively for sequencing duties (so that it would start to feel more natural and I could get away with less reliance on the manual or cheatsheet). It’s worked pretty well. There’s also something about the system that feels more like a unit without the external control tools (acknowledging there is the computer keyboard attached, that “feels” slightly different that a control surface to me though, I think because I’m not typically changing my scripts a ton after I have set the patch up, and rather exploring them by altering triggers to those scripts, most often by my tempi)
I think one more thing to take into account when thinking about the modular as an instrument is the ability to recall. You kind of get to this when you mention:
Focusing on the Teletype has taught me I really value having recallable sequences (or rather, melodic sequence “systems” that return something vaguely melodically cohesive, even if things aren’t like 100% set and have some level of randomization). I think the melodic content, for my own music, is second (or maybe rather, I feel like I like it to exist in a different creative phase) to the timbral and changing sounds over time that I really enjoy spending my time exploring.
Another thing I’ve found is that I personally don’t have the patience/confidence to come back to something simple and build it into something more complex. I feel much more comfortable and in a creative space if I can more-or-less have a finished thing that I record (using effects afterwards to make it more full is fine) So for me, going the modules-that-make-up-a-monosynth route would probably be a ton of fun, but I think I would have a lot of trouble utilizing the output in some sort of “finished” work if it needed several passes.
And one more thing that I really value are modules that have many uses. I feel like I have some pretty core flows that at this point I reach for, but it’s fun to have the potential to explore wildly different flows (cause somtimes just creating something new on something I feel pretty confident with is not super interesting). Just Friends (can be a sound generator, trigger sequencer, envelope generator), Rings (can be a simple poly-ish voice or a really interesting effect that reacts to inputs in sometimes pretty unexpected, non-linear ways), and Teletype are all good examples of this for me.