That's the point - you don't start a new project every 2 weeks. You start off with an idea easily adaptable to various scenarios. Then adapt each time, target n

That's the point - you don't start a new project every 2 weeks. You start off with an idea easily adaptable to various scenarios. Then adapt each time, target new people, and observe if the variation makes you required $1 a day. When you have nichified a codebase that nothing more can be squeezed of it, you start on the next.