Are perfect mechanics an achievable objective in training? I would propose that you must have a concept or an image of what “perfect mechanics” would be for the movement you are preparing for, but to achieve perfection in terms of mechanics is at best a moving target. It depends on how you define perfect. I believe that in each activity there are acceptable ranges that vary from person to person. Certainly we can create computer optimization models based on algorithms derived from biomechanical analysis but the human body is not a computer, there are too many variables to control. It is best to have a sound technical model and work to adapt that model so that it fits the individuals that you are coaching. In developing technique the danger that I have see is the robotization that occurs when someone tries to impose the perfect model on an athlete. This is usually done through detailed step by step progressions. There is nothing wrong with progression, it is essential, but be aware of paint by numbers progressions that break the movement down into too many small details that do not connect to the whole. For example one of my pet peeves in teaching throwing is to start someone throwing from their knees. Take a step back and look at it, you are taking out the biggest link in the kinetic chain and adding a disconnected skill. How do you learn to use your legs if you kneeling?
I am convinced the best approach to technique teaching and development is to give the athletes movement problems to solve, see how the solve it and keep tweaking and modifying until they can replicate the movement with reasonable resemblance to the technical model. Each athlete has a movement fingerprint or signature; they will find that, through discovery. This approach to training will be a mindful experience for them. It comes back to a recurring theme in this blog – train to create athletes that are adaptable rather than adapted.