I'm a big fan of alternatives, but have come to realize there often valid reasons why they aren't used.
We have several systems for matching resources to production because there isn't any single way that works best. But it is interesting to try to stretch each model as far as it can, for best results we can get.
Can you elaborate on what you mean by "risky"? I can only think of projects that don't yet have a customer. In that case, you just don't build it until there is a customer need for it. Architects don't speculatively design buildings hoping to find a customer who wants it. I think we would be better off if the same was true of software.
I am curious if there is a open source type model for getting a higher return for doing riskier work, as economics would suggest is fair and also helpful.
That is obviously a typical closed source situation, but people find creative solutions.
Not talking about selling services, but somehow compensating for writing the code after the fact, risk adjusted, while keeping the code itself free. Might need a temporarily hybrid license to accomplish this. I.e. the code is obligated to become free, after some fixed license fees have covered the work cost and risk.
This is what labor contracts are for, no? I fulfill my side of the contract (by providing software) and then you fulfill your side of the contact (by providing money), else if I fulfill my side and you do not fulfill yours, I have a legal case against you.