I work in a related field (digital logic for embedded systems) and identify with the description you gave. I find it's best to do analysis on the potential bene

I work in a related field (digital logic for embedded systems) and identify with the description you gave. I find it's best to do analysis on the potential benefits/drawbacks and prioritize. Over time, you get a bunch of ideas and the pet ones fall to the wayside as it's clear that there are other features you could be working on that provide better value for less effort.