> In stressful situations I always have to fall back to exploitation-only, possibly getting stuck in the local minimum of what I'm doing.
The situations where you are with your back to the wall and have to hunker down to meet a hopeless deadline are exactly the situations where exploring solutions benefits the most. Taking 5 or 10 minutes before you start down a path to consider the problem from the outside does not hurt the time to completion, and often benefits it when there are relatively obvious alternate solutions that are more narrowly scoped.
Spending longer than a trivial time on exploration is only really worth it when the possible upside is also very big. Before starting a year-long project I would find it normal to spend a few weeks exploring ideas, instead of just hunkering down to write code and seeing where it ends up.
>> Taking 5 or 10 minutes before you start down a path to consider the problem from the outside does not hurt the time to completion
This is true for modern, civilized world. We humans still suffer from "savannah brain", where main cause of stress was physical danger, and taking even 5-10 seconds to consider your options might result in you ending up dead.
That's why we developed physiological reaction to stress, that with help of hormones such as adrenaline and cortisol shuts down exploratory part of our brain on chemical level.
Following your suggestion of taking 5-10 minutes to reconsider requires serious effort to overcome that reaction and, depending on your individual physiology, might even be not possible at all.
On top of that there's the thing called "stress addiction", which basically means that some people start seeking new sources of stress when their live becomes to peaceful. Which again works against your exploratory brain potential.
The situations where you are with your back to the wall and have to hunker down to meet a hopeless deadline are exactly the situations where exploring solutions benefits the most. Taking 5 or 10 minutes before you start down a path to consider the problem from the outside does not hurt the time to completion, and often benefits it when there are relatively obvious alternate solutions that are more narrowly scoped.
Spending longer than a trivial time on exploration is only really worth it when the possible upside is also very big. Before starting a year-long project I would find it normal to spend a few weeks exploring ideas, instead of just hunkering down to write code and seeing where it ends up.