This approach is, unfortunately, not going to fare too well in reality. It does work well as long as every single individual in the entire process is perfectly omniscient, though!
One of the major issues here is the design documents. If you look at your process, the key fundamental objects being passed around are information. Your idea person generates a sales pitch which is distributed to the internet. They also generate a vision statement for the idea which the professional designer uses as a seed to create their design. This design, of course, is itself just information.
Now these documents, in reality, are never perfect. I've invested in several crowdfunded ventures and been terribly disappointed with the final product. Had I not been an early funder, the final product would certainly never have caught my eye. So we're going to have to deal with the inefficiency of people not getting what they thought they were going to get. Indeed, it is worth noting that I didn't call this design document "the idea of a product or service," mirroring the terms you used. I called it a "sales pitch." The most successful pitches in your model are not the ones which are good ideas. They are the ideas which are sold well, good or bad.
In the non-crowdfunded world, we have to deal with this process as well. It's easy to have an idea that doesn't quite live up to its buzz. However, we have entire systems with dozens or hundreds of people involved to create a more robust process that produces the required product more reliably.
Also worth noting, the idea itself isn't all that valuable. I like Howard Tayler's take on it. I like it so much that I think it wise to reproduce it here:
Anytime people ask me where I get my ideas (and it happens all the
time) I immediately jump up on a soapbox and explain to them that
they're asking the wrong question.
My ideas, your ideas, and everybody's ideas have no intrinsic value,
so it doesn't matter where I get them. They are not currency, they
cannot be bought or sold, they are, in market terms, worthless.
Worthless, that is, until somebody does something with them.
Right now I have several ideas percolating for scripts I should be
writing. Those ideas have no value until I actually write the scripts.
And in truth, the scripts themselves have limited value until I draw
on them, and Travis colors them, and we put them on the web.
Eventually those ideas become a product that I can sell, and then they
The only monetizable product from this ideas step is a sales pitch, which ropes people into putting down a deposit on a product they've never seen, relying only on the words of a salesman.
We then get to the professional designer's work. You use the singular here, but for all practical purposes worth considering, the design work is a group activity. Even crowdfunded tasks typically have a team of 5-10 people. This design effort is no small thing. Rarely is one designer sufficient. For real interesting products, the teams are even larger. The team that turned "We want a CPU that has great branch prediction and a very fast L2 cache" into the Intel Core series of processors was hundreds if not thousands of designers!
And you stick them with the risk. The risk is on your designers if the product doesn't go to market. Accordingly, they will need to seek profit. There's no way they can do the design "at cost" because they have to cover the risk of the product flopping.
The designers are also not necessarily "in" on the idea. All they see are the design documents provided by the person who had the idea. In this sense, I'm reminded of a contest I saw as part of Odyssey of the Mind many years ago. The challenge was to build a lego structure as a team of two. The trick? One person gets to see the goal design, and must write down instructions for the second person to follow. Those instructions are then acted upon by the second person, who is not allowed to ask any questions. The challenge is to build the lego structure as quickly as possible (corresponding to as cheaply as possible in your system), without mistakes.
It's not easy.
Myself, I work on software. 100% of the user-provided "ideas" that I implemented required substantial iterative development to massage it from ideas-speak into a form which is more amicable to design. Most ideas I was given I rejected outright, simply because there was no affordable way to implement anything resembling what they wanted, even though they thought the idea should be cheap.
By the way, did you notice that your process involved people putting down deposits for what they think the idea is worth before a professional ever looked at the design and told you how feasible it is?
Now we get to the computer. The computer is responsible for the factory. Managing a factory is hard. It requires a tremendous amount of skill. And it's not as straight forward as you suggest where the computer just decides prices and quantities.
Most products can be designed several ways. In small volumes, one may 3d print a design. This is very costly, tying up an expensive machine for a very long period of time. In larger volumes, one may invest in "tooling," and construct molds which permit constructing these designs en masse. In even larger volumes, one may invest in entire toolchains custom tailored to meeting your needs.
Each of these require not only massively different levels of initial investment, but also require design changes. You design products differently depending on what their manufacturing process will be.
So the computer also has to make predictions of how many people will want, based on ... something. We know the initial deposits are a bad source of data because those people never got to see the final product, nor how well it performed.
We usually have entire teams of people, operating early in design, exploring market sizes and identifying the best manufacturing methods. These people are using computers, so they will always have an advantage over just a computer on its own.
Also, there is a question of R&D. I have a great idea: "lets make a car that earns 2 MPG more than the current model." A bunch of people buy into it, because MPG are good for the environment and the pocketbook. Now how do we make that a reality? Believe me when I say design firms spend millions of dollars trying to find ways to make those ideas into a reality. If they succeed, they like to keep the earned intellectual capital for themselves. This feeds quickly into the risk that the professional developers take.
In the end, consider your average design firm. Pick any one you want. How much attention do they give to unsolicited suggestions? It's highly unusual for such firms to take such suggestions (so unusual that I can't think of a single well known example of it happening). If this economic model was effective, it would suggest that design firms have a tremendous financial interest in getting unsolicited ideas (with a few deposits backing them).