It's a bit of a mental shift. Over 100-150ms you're listening for delay - it's the same skill that lets you identify a tempo. You're trying to listen for the gaps between sounds.

Once you get below that range, you're no longer listening for a delay, you're listening for thickness or width or whatever you identify it as. It's more like an audible biggening than anything, it's hard to describe.

Above that range you can get by without using the on/off switch. At that range (at least in the beginning) it gets super hard to tell without toggling the button. The game at those ranges stops leaning toward being Metronomer and starts leaning toward being Stereohead.

@Jay Kay that's such a good explanation; as I've gotten into music theory and tried to learn subdivisions, I've noticed that delay is typically dialed in specific ways, e.g., Neil Young on his solos seems to like the strum/pick plus delays to equal 3 over 2 beats, and other folks may be doing any number of repeats per beat. The question is, do many musicians actually know they're doing this, or is it mostly by feel? I'm guessing it's mostly by feel, especially live, but I imagine producers are more likely to be exacting. But like you said, it's so interesting how perceptually, the countable nature of the repeats disappears as the time shortens, and instead you're left with that thickness or width, and how you determine what you want is based on feel.