That whole signal surfacing out of subspace thing beyond 22 ly or so does present some problems - either instant comms within range of the network or lightspeed only to the nearest relay station once the signals surface.
We probably have to assume there's some fuzziness close to the dropoff point, so that the message first slows down slightly to account for all those instances of couple of hours or days of delay, and only then degrades to STL...
Then again, signal delay is a rare event in Star Trek: it's only really witnessed in "The
Enterprise Incident" and "The Defector" IIRC. There is some delay in establishing communications in, say, "Ensigns of Command", but after the initial fuss the comms then become instantaneous. Probably a matter of interfacing a number of alien relay networks with the Starfleet/Federation one (such networks put all together should cover most of the galaxy by now, one would think).
So perhaps the dropoff really is sharp as a knife, and the delays in those episodes are rare coincidences. Or not really coincidences, but rather the result of Romulan jamming or Romulan search-and-destroy-comm-relays sorties in the vicinity of their Neutral Zone.
I can't remember if starfleet actually received some sort of transmission from "Friendship One" in the Delta Quadrant or not in the related VOY episode, but I'd like to assume that the surfacing phenomenon is inversely related to signal propagation speed, which is to say that slower signals might stay "submerged" in subspace over much greater distances than 22 ly.
Might be that. But it might also simply be that the more powerful signal stays up there longer, and the interstellar insta-signals we usually get are powered by planetary arrays much more potent than a
Galaxy class ship's transmitter. It just suffices to feed that signal strength to the system at some point, not necessarily at every relay and endpoint. (It would take some Treknobabble to explain how that's possible, but still...)
Friendship 1 did seem to remain in constant contact with home base, as Janeway had this to say:
Janeway: "In any case, we lost contact with the probe one hundred and thirty years ago but, its last known co-ordinates..."
That is, the probe was launched in 2067, and lost in the 2240s, at a time when it was already in the Beta/Delta Quadrant border. Its location was known with pinpoint accuracy - all
Voyager had to do was fly a sublight search grid. That would seem to imply fast communications across vast distances. Granted, news of the loss might have taken decades to reach Earth, and the 2240s date would have become known only when the last transmission faded out sometime in the 24th century. But the loss was already known at least in Harry Kim's childhood, so the delay would be at most one century across some 30,000 ly. And that's without a working relay chain for most of the distance.
Also remember how Data seemed to think it would take mere three centuries to send a message across intergalactic distances in "Where No One Has Gone Before". Perhaps with just a little extra oomph to the signal, you get much greater persistence at subspace, and very, very fast communications across immense ranges. But if the
Galaxy really has the ability to produce that extra oomph in a cinch, why the standard limitation of 22.65 ly?
Perhaps it's a matter of bandwidth? A "Hi!" can carry across hundreds of thousands of lightyears, but a "Hi there!" peters out at mere thousands, and a meaningful data burst barely clears the first few dozen lightyears before degrading. Subspace telescopes could then see across vast ranges by utilizing very simple pulses, but those pulses would not do much good as a means of communication. And the E-D could send a special "distress pulse" from one galaxy to another, but would not be able to relay back her findings or ask for instructions.
Timo Saloniemi