Because of the freedoms given by GPL, there is a problem that competing GPL marketing entities will compete themselves to death, since they could always take the code (that they don't pay to develop) and sell for near zero cost that doesn't recoup the development cost.
I think I'm missing a point here. What would the other marketing entities sell for "near zero cost" ? I see nothing in the Free Software that can be sold for near zero cost. If the marketing entity sells boxes (Debian distribution for instance) including software they did not pay to develop, they still have to pay for making the packaging and it's not "near zero cost". In short I have trouble to see why competing marketing entities could not survive. Sure, they will compete but what's wrong with that ? Can you explain ?
I was being unclear. What I should have said is "near zero development cost,"
that won't fund the developers work, since they can just take the code under the GPL for zero cost. They will still have variable marketing costs, that they each must recoup. And depending on how much friction the market has, some marketers can sell for more than their variable marketing costs and maybe even make some profit.
But the main issue is the wrong people are getting the money. Most developers will get nothing, because for any one marketing company that tries to charge more so it can fund development, other marketing companies will sell the same thing without the development cost component. Therefore, there is a cost problem in doing the right thing and paying the developers, that competing marketing firms won't have.
That is why a superstructure is needed to incorporate the development cost into the equation. Once that superstructure is in place, then competing marketing firms can compete and fund the development at the same time, which is the independent consultant layer in the CommCo. The FSMC becomes a wholesaler and the others are retailers, so that the development costs are funded.
The development cost in free software is currently an externality that must be brought into the equation for a properly functioning economic solution. This is like pollution made in one country and causing problems in another. The solution is an international superstructure to coordinate the national bodies, so that the full cost is incorporated in the right entities. Otherwise the economic solution is suboptimal.
You also need to remember that free software must compete with proprietary software, not just with other free software marketing companies. And without paying for the development in free software, proprietary will have most of the developers, because it pays its developers, even though proprietary is less efficient as a development method. Without incorporating free software development costs into the pricing to fund it, proprietary as a community will always have an advantage over free software, because most developers will code for it and resist the free paradigm.
Now, free software development is so much more efficient that it is doing quite well against proprietary development even with this proprietary advantage. But the real question is how much more great software would be created and faster, if everyone worked in the free development model. But to get the 4 million proprietary developers to work in free, they must be paid.
Once free software pays its developers, then it not only has the efficiency of free development, but will also have the former proprietary developers. Also, most free developers work somewhere else to pay the bills. So, if they get paid to develop free software, they will produce more free software, because now they don't have to work somewhere else to personally fund their own free software development.
These efficiencies all interrelate and multiple, because all the obstacles and countervailing forces are removed.