In my opinion, for most projects what's necessary is to approach from a broadly informed perspective and then dive deep into multiple potential solution strateg

In my opinion, for most projects what's necessary is to approach from a broadly informed perspective and then dive deep into multiple potential solution strategies to see what direction will work well. Then dive even deeper into the details related to the architecture you have chosen. The problem with focusing too much on one area is that you can only deal with one particular type of problem and maybe only with one particular type of solution.