I think you are all right here. In my opinion you indeed need a prototype phase with a strong focus on gameplay. But at the same time you should have a clear vision of style and mood you want to create. Arts, music and story are very important to achieve all this. In the end you will fail if only one of these components are too bad to please your potential customer.

And that is the reason why small indies have small projects: The successful ones make good games with good gameplay, good graphics, good music and an interesting story. And to finish this in time they choose something smaller. All the successful stories I have read about started quite small with casual games, flash games, web or mobile games, adventures or simulations. After that they grow bigger.

And they all tried to balance all components of their products, not only game-play. It would not work at all, just take Adventures as an example. The gameplay is not new at all, the story, the art and the mood will immerse the player.

A gamer expects that gameplay just works, that there are incentives to go on, that there is something to explore, something interesting to reveal. I cant remember any gamer to say: "Hey, the controls were amazing and the camera view was so great!"
While this is really very important especially in a prototype phase, but the gamer wants that it just works and will probably not mention it as a feature of your game, except it is really very exotic.

I think people mix up the priorities for developing vs finishing-up a game.

During the early development, the gameplay is key.
Thats the prototyping phase.
There indeed its more distracting to use complex art assets
and content before the core-gameplay and user-controlls are
tested. (at this phase a working gamemechanic must be created,
a functional prototype is a must. Cranking out Art already slows the progress and fexibility of this phase.)

But when finishing up the game, creating the content and Art assets are key.
At this point the gameplay should be set and tested.
The Artists know how the properties of the assets must be.
The gameplay is not changed, but the leveldesigners can
use variations of the provided core-mechanics.

In the end its about polishing. This phase is the most straining one.
Iterating though the game again and again to smooth out flaws.

------

So when I complain about a "new super" project that shows
a super high-res model in a standard-texture hollow-cube room,
its not about sayning it should not have nice Arts.
Its about pointing out, that this project probably does not
even have a playable prototype version that covers most
gamemechanics.
Its therefore not even a game yet. Just an Art display.

Yes, I absolutely agree with you. This was what I wanted to describe but you brought it much better to a point. A prototype is about gameplay, mechanics and to test how it feels. Some blocky place-holders are sufficient there.

But I also would create an art-prototype at the same time. It can be a set of concept drawings but it helps to find the right direction right at the beginning and can be a good set of documents to instruct all later involved artists to stick to the same style.

Shure, a mockup scene helps to visualize game in development.
So if Art is already there (eg these content packs), it
can help to determine the style direction of the game.

But this mockup should be kept seperate (just a stage) from the actual game-code
until the game-architecture has matured enough to implement
the art-assets propperly.

(Hobby)Developers should not build a game around a Gun or Soldiermodel.