When I say “socialize the costs,” I mean that the burden of supporting/paying developers can be put on the entire community from the limited/shared pool, even if a super-majority of stakeholders don’t want the development that would be funded.
When grants of money can be given so easily to developers with little to no risk to any individual voter, chances are that accountability will be lacking for both the fund allocators and the developers. After all, it’s not their money that’s being spent. It’s the same problem we see with government and tax dollars. Are politicians rushing to give their own money to the state projects that they so proudly support with their rhetoric? Of course not. But they’ll certainly use that shared pool of cash that didn’t come out of their bank accounts...and they have no reason to make sure it’s well-spent either.
I know software development isn’t cheap. I know developers want to be paid for their work. But I also think that, if the proposed development from a trusted person/entity is truly valued by investors, then those investors will find a way to fund it and the developers should be able to solicit funding for it.
I know this is fairly new tech and some new concepts when it comes to organizing and developing, but people and businesses still need to put in the leg-work and not just hope for token giveaways.
But that conversation is practically irrelevant and probably a distraction, so I’ll just say this...
As long as there isn’t an increase in the inflation rate, I really don’t have much of a problem with a small percentage of the overall rewards pool going to development. I just hope that these funds are more carefully considered and allocated than post rewards and that the vetting of developers is better than that of our witnesses.
I would have serious concerns about large accounts (like the soon-to-be-anonymous Steemit accounts) directing funds to their own “projects” that are disguised as “community development.”
But ultimately, I would be more in favor of a development fund that operates on actual donations rather than inflation-funded reward pools. If there aren’t enough people willing to support useful development, then we have a real problem.
Just to briefly address your first point: if a majority of the stakeholders don't want a proposal to be funded, they can always vote the refund worker proposal in front of that proposal and the undesirable proposal won't be funded.
In fact, this ability to stop all funding actually upset a lot of people in BitShares in the distant past, because some large stakeholders decided they didn't want to fund any more work by the core team at that time and did exactly that.