Alex Iskold of Read/Write Web has an a perspective worth reading as looks at the inner workings of Yahoo! Pipes. Reading his opinions, I think he may well be on to something, but perhaps the steep growth 5000 pipes he discusses will likely decelerate n momentum until either the infrastructure can't take the load or we run short on ideas. Reapplying established relational database thinking and methodologies on this new paradigm seems like an obvious lateral shift, but as it's stands I think it is an interesting academic area of exploration, and too green for any serious corporate investment. Whether it will accelerate the path to building mission critical applications in this manner, consider the years it took to deliver effective and dependable applications using relational database. This model presents a new set of challenges in terms of infrastructure, reliability and performance; all of which need to be overcome before this model can find it's place as a critical application framework.