This is so inspiring, thanks a lot. It's great to see that there are now resources going into supporting OSS developers that require only an idea and pedigree, and not a finished product and a business plan.
Innovation is NIH somewhat, somehow and that is a Good Thing.
I hate working at a place where people are so busy being clever that they reinvent everything. You steal from the younger devs an opportunity to learn, and you externalize the cost of your poor documentation and interface consistency onto your coworkers instead of spreading it out over the entire industry. This is NIH.
What you should do is pick a segment of your problem space that has mediocre tools and make something much better. The 80/20 Rule seems to work quite well here, when I see it applied, which is not often enough.
That is 20% of the effort can deliver 80% of the features?