I agree with this. I also did this for years. Then a couple of years ago I buckled down, built an MVP (mininamal viable product) in the most brain dead way possible, and put it out there. And... nothing, my idea received zero interest.

So the really cruel thing here is, all the dithering and indecisiveness really do amount to nothing, none of the decisions you mentioned matter; language, framework, db, etc. If something you build does gain users, then worry about it!