sometimes a person seemingly doing unrelated things is still 'working'. i know at least for myself, i run through various scenarios of the outcome rather than staring at the code before i make my decision to implement.
edit: and i'll typically run through those scenarios while doing seemingly unrelated stuff, like shitposting on reddit or sitting in a blacked out room.
In a similar vein I often like to try out a few avenues, because thinking about them won't get me any further, i need to interact with them to build meaningful scenarios.
Of course if you try three ways to get the right way, you've "wasted" time on trying to two wrong things.
E.g. when selecting a library it's often hard to find the one case that breaks, just from reading documentation, instead playing around with the library can get you there faster.
edit: and i'll typically run through those scenarios while doing seemingly unrelated stuff, like shitposting on reddit or sitting in a blacked out room.