The lowest levels of the stack are too ossified to have your ideas adopted, while the highest levels are already idea saturated. The most interesting place to innovate is a thin slice in the middle, where small changes can have radical effects. Right now I believe the sweet spot is identity and data storage of small payloads