The essence of any good design starts with clearly defined specs on what you need.
Torque range, speed range, inertia or load, cost, R&D effort & time, power source options.
Then step load response starting and applied load. RPM vs step torque
From this you can select power, budget, engine and a control system, then controller, feedback and detailed design.
Anything else I missed can be added. Without this systematic approach, you will have too many surprises, which are inevitable. Murphy's Law.