If you're going to publish a map of your imaginary land, map out the watersheds and make sure they make sense. You don't have to include those details in the final map, but they should inform the final map.
Real rivers go somewhere. They don't just stop in the middle of nowhere.
My opinion: Features like this have no business being in a fantasy map unless the author has an above-average grasp of geology.
Or there's a magic portal somewhere. That's fine, just don't pretend it's natural.
I'm complaining about fantasy rivers that literally just stop.
Actually, ignoring the issue of whether Middle Earth was geologically feasible, Tolkien went to great lengths to make sure that the story was consistent with the geography, e.g. in terms of making sure character times and locations were synchronised across the many plot strands. The LOTR map was specifically created to help him keep track of it.
[0] https://www.tor.com/2017/08/01/tolkiens-map-and-the-messed-u...
The creator, Eru / Ilúvatar creates a great theme and the Ainur (spirits) are then invited to create a music from it. The evil Morgoth (Sauron's boss) sows discord by breaking Eru's harmony. Eru then stops the music and the various spirits manifest in the world (Arda) created from it. The world is initially illuminated by stars and then the light from two great trees. Which Morgoth destroys, although their light bearing fruit are used to create the Sun and the Moon. The Valar (the most powerful good Ainur) create the westerly island Valinor, which (after a lot of misery which forms the main subject of the Silmarillion) Ilúvatar eventually makes inaccessible by sending a great wave, changing the shape of all the lands of Middle-earth.
Hence my comment that plate tectonics (or whatever geological processes were understood in the early 20th C) wasn't an overarching concern for Tolkien. But once he'd made this world, he made sure that the LOTR characters would see the same phase of the moon at their respective parts of the story.
Anything that deviates from these rules is usually wrong. Rivers don't split around two sides of a mountain, for example.
The Orinoco River and Rio Negro are connected via a distributary, although this is the upper reaches of both rivers.
Deltas are commonly very large and involve several streams. The mouth of the Amazon River is more or less 100 miles distant. And the Brahamaputra and Ganges river delta involves 100s of channels that more or less crisscoss the entirety of Bangladesh.
The author is saying two streams travelling up never split, only merge. I'm saying they do split [0]. An upstream split is a downstream merge.
Yeah? Have a look at the Morghab River in southern Turkmenistan. It flows from some mountains in Afghanistan out into a desert, where it splits up into a delta and just sort of... stops. The satellite view on Google Maps is striking. (There are several canals connecting to the delta now, but historically the river stopped in the middle of nowhere.)
Rules of thumb like this aren't laws, just patterns which have exceptions. (That said, yes, imaginary maps are sometimes ridiculous.)