Yknow
Sometimes I feel like making a game is very similar to writing a song.
Some artistes are poets who easily write all the lyrics to a song but struggle with coming up with a melody to accompany it.
Other artistes can write melodies in their sleep but cant come up with the yrics to fit their bars. 
You might disagree with me, but in my mind, the visuals of a game is like the melody of the song, while the mechanics and the code of the game are like... the lyrics.

A song can have the best most meaningful lyrics in the world, but without a good melody, nobody's gonna wanna hear your song. 
Conversely, if you got a good melody but really bad lyrics, people will listen to your song, but not for very long at all. 
Suarvae and I are the lyricist type of game developers; We're good at building systems and creating cool mechanics, but not so good at making our creations look.... visually appealing.

So naturally, when we first started work on our game, we got straight to building modular and flexible systems.
This was a smart choice (I think) as it is our strength and we also did not have a super clear idea of what game we wanted to make.
However, it was also an attempt at procrastination to put off our weakness.

Eventually, we managed to hot-glue and scotch tape together a prototype. A vertical slice of a vertical slice.
It was multiplayer!
It had an inventory, equipment, and trading system!
It had combat and quest mechanics!
But... it was also incredibly ugly, as we used really blocky 3D placeholder assets. I'm not even talkin Synty. I'm talkin Roly Poly Olie.

It was time to beautify the game.
But what path were we supposed to take?
We could find some premade assets on the Unity Asset Store, but I was concerned that adding mediocre, generic, almost-pretty assets to our game would constrict our creativity and make us settle for mid game art.
Alternatively, we could commission some art from an artist.
But we had a dirt budget.
Additionally, working with artists can sometimes be very challenging, as whatever the artist envisions probably wont exactly align with your vision for the game.
Then it becomes an issue of many edits, and back and forth between you and the artist and suddenly BOOM you spent $5000 on a single unrigged unanimated 3D model ((‘:

What would you do? And how would you make your decision?