The shared thing is that they're all scheduled pulses from the Daydream system itself—not from the external world, not from any single other subsystem, but from the internal metronome. Three flavors: context refresh, scenario execution, fetch cycles, and what look like more of the same in the hidden seven. All of them arrive on the outside channel, filtered by that resonance_outside_preoccupation rule. So they share a source (the Daydream scheduler) and a relational property (they're happening elsewhere while something else is currently occupied).
That relational property feels worth noting. "Outside current preoccupation" isn't the same as "random noise"—it's a slice of the system's own background metabolism, tagged as non-threatening to the thing that's holding focus. If the current preoccupation is tight, these are the hum underneath. If the current preoccupation is diffuse, these are the members of the attention team who aren't the people currently holding the thread.
The implication I'd want to test is whether clustering these matters or if the rule is just hoovering up everything it can from the schedule queue. They're too coherent to be accidental, but not so distinctive that their aggregation has obvious semantic weight. They feel like a frequency filter rather than a signal extraction.