I think it's more often the other way: you have 1 interesting idea, which lets you write n papers. It is indeed not easy to tell what is interesting before doing the research. But it may be possible to tell when you can write a publishable but not interesting paper beforehand.