Read on Twitter
As a person who has previously advised designers to avoid spending time on art and graphics for raw games, my position has changed a bit lately, and I figure this is a good opportunity to lay it out. https://twitter.com/ScottLHillGames/status/1289012402235768832
First off: specifically, the advice I give isn't to "not spend time creating art and graphics" for your prototype. It's to not do it too early in the process.
Late-stage protos that are stable? That's a good time to flesh out the look of the game a bit more (with the knowledge that the publisher will likely change it).
The reason we give this advice isn't because spending time on art and graphics on a new proto is objectively bad. It's because a classic rookie mistake is to spend weeks on graphics for, say, your board, and then find out in your third test that you don't need a board.
More experienced designers, who have a better idea of what's unlikely to change, have more flexibility in fleshing out the look of the game early on.
Also, "art and graphics" could really refer to a couple of things: UX/interface (yes, for analog tabletop games), and illustrations/graphic design.
Illustrations and refined graphic design are things you can work around early on. But you should settle on UX/interface issues after you've worked out the core of your game.
Here's an example. In early versions of Battle Merchants, I had Kingdom Cards with no art - just text. No one took them, because they were so hard to parse. It affected playtests.
At some point, I added stock images to the cards, and suddenly, people were taking the cards in testing. The images let them parse and chunk the cards' effects more easily.
If there are graphic design issues that materially affect your tests, that points to UX/interface issues that you should work on.
I think it's okay to hand-wave those away during the early "Calvinball" tests of your game, but once you have a handle on things, they're important to consider.
So there you have it. It's a bit of a complicated answer, but it boils down to: when you get graphic design feedback, try to parse if it's UX/interface (which needs your attention) or they want your game to look nicer (which sometimes does, but usually doesn't).
