Yep. Over the last two years i've been fortunate enough to have a lot of time on my hands. In that time i've built 3 and a half successful products. 6 days of (thinking mostly, some research) about what to do. 1 day of actually doing it. It's so easy to spend that one day completely and productively - I already know what i'm supposed to do, I just have to sit down and do it.

The 6 days (not always 6 - maybe more, maybe less) let me come up with and throw out stupid ideas before i've coded them. By the time I come to build something it's a carefully considered thing that'll actually be useful most of the time.

It's not just feature ideas either, it's code design, data design, anything that might be complicated enough that I used to spend multiple days experimenting. It doesn't always work out, but when it doesn't, i've thought it through enough that I have a head start on coming up with a better design.