I think most engineers need to be both. In most situations at least, you are deciding on engineering time vs non-engineering time. Meaning, if I can build something in a day that "gets the job done" with some usage constraints, or I can spend a week to build something that removes all constraints, then I need to calculate the trade-off. Could something more useful be built with that extra time? At least in start-up world, I find myself calculating that trade-off a lot.