While a great idea is compelling, a far more important factor is who is carrying out the work.
A US government study in the (1960s?) recommended that the most effective way of funding research would be to find the best researchers, fund them, and let them do what they want.[verification needed]
Great projects are rarely if ever run by terrible leaders. Competent international development work can be done by bad leaders, but not great work. It's the same in companies - no amount of quality systems can make up for a lack of good leadership at the top.
But we give our grants and contracts in development work based on the structure of the project, as though the logframe and the staff chart are all that matter. This is a problem. It's not the shape of the staff chart that matters, it's the names you find in the chart.
We should give money to people who can do good things with it. There are brilliant people aid and development, as well as in business, who have the ability to do extremely effective things with their funding - but often don't get adequate funding, or spend much of their time looking for funding, rather than being able to apply their genius full-time.
Instead of supporting people who can get things done, we support structures. Some of that is an effort to support fragile states and build democratic institutions, and we should keep it up. Some of the support for structures, however, is just intellectual laziness, and a desire to do things how they have always been done. That needs to change. This sounds like a recipe for corruption, but anything is a recipe for corruption if you do it badly. We can still ask for financial accountability, and for proof of results.
Innovation comes from individual people with ideas and passion. We should find and support those people to bring change in their countries and communities.